Network Design, Modelling and Performance Evaluation-The Institution of Engineering and Technol

IET TELECOMMUNICATIONS SERIES 77
Network Design, Modeling,

and Performance
Evaluation
Other volumes in this series:
Volume 9 Phase Noise in Signal Sources W.P. Robins

Volume 12 Spread Spectrum in Communications R. Skaug and J.F. Hjelmstad
Volume 13 Advanced Signal Processing D.J. Creasey (Editor)
Volume 19 Telecommunications Traffic, Tariffs and Costs R.E. Farr
Volume 20 An Introduction to Satellite Communications D.I. Dalgleish
Volume 26 Common-Channel Signalling R.J. Manterfield
Volume 28 Very Small Aperture Terminals (VSATs) J.L. Everett (Editor)
Volume 29 ATM: The broadband telecommunications solution L.G. Cuthbert and J.C. Sapanel
Volume 31 Data Communications and Networks, 3rd Edition R.L. Brewster (Editor)
Volume 32 Analogue Optical Fibre Communications B. Wilson, Z. Ghassemlooy, and I.Z. Darwazeh
(Editors)
Volume 33 Modern Personal Radio Systems R.C.V. Macario (Editor)
Volume 34 Digital Broadcasting P. Dambacher
Volume 35 Principles of Performance Engineering for Telecommunication and Information
Systems M. Ghanbari, C.J. Hughes, M.C. Sinclair, and J.P. Eade
Volume 36 Telecommunication Networks, 2nd Edition J.E. Flood (Editor)
Volume 37 Optical Communication Receiver Design S.B. Alexander
Volume 38 Satellite Communication Systems, 3rd Edition B.G. Evans (Editor)
Volume 40 Spread Spectrum in Mobile Communication O. Berg, T. Berg, J.F. Hjelmstad, S. Haavik,
and R. Skaug
Volume 41 World Telecommunications Economics J.J. Wheatley
Volume 43 Telecommunications Signalling R.J. Manterfield
Volume 44 Digital Signal Filtering, Analysis and Restoration J. Jan
Volume 45 Radio Spectrum Management, 2nd Edition D.J. Withers
Volume 46 Intelligent Networks: Principles and applications J.R. Anderson
Volume 47 Local Access Network Technologies P. France
Volume 48 Telecommunications Quality of Service Management A.P. Oodan (Editor)
Volume 49 Standard Codecs: Image compression to advanced video coding M. Ghanbari
Volume 50 Telecommunications Regulation J. Buckley
Volume 51 Security for Mobility C. Mitchell (Editor)
Volume 52 Understanding Telecommunications Networks A. Valdar
Volume 53 Video Compression Systems: From first principles to concatenated codecs A. Bock
Volume 54 Standard Codecs: Image compression to advanced video coding, 3rd edition
M. Ghanbari
Volume 59 Dynamic Ad Hoc Networks H. Rashvand and H. Chao (Editors)
Volume 60 Understanding Telecommunications Business A. Valdar and I. Morfett
Volume 65 Advances in Body-Centric Wireless Communication: Applications and state-of-the-
art Q.H. Abbasi, M.U. Rehman, K. Qaraqe, and A. Alomainy (Editors)
Volume 67 Managing the Internet of Things: Architectures, theories and applications J. Huang
and K. Hua (Editors)
Volume 68 Advanced Relay Technologies in Next Generation Wireless Communications
I. Krikidis and G. Zheng
Volume 69 5G Wireless Technologies A. Alexiou (Editor)
Volume 70 Cloud and Fog Computing in 5G Mobile Networks E. Markakis, G. Mastorakis,
C.X. Mavromoustakis, and E. Pallis (Editors)
Volume 71 Understanding Telecommunications Networks, 2nd edition A. Valdar
Volume 72 Introduction to Digital Wireless Communications Hong-Chuan Yang
Volume 73 Network as a Service for Next Generation Internet Q. Duan and S. Wang (Editors)
Volume 74 Access, Fronthaul and Backhaul Networks for 5G & Beyond M.A. Imran, S.A.R. Zaidi,
and M.Z. Shakir (Editors)
Volume 76 Trusted Communications with Physical Layer Security for 5G and Beyond
T.Q. Duong, X. Zhou, and H.V. Poor (Editors)
Volume 79 Satellite Communications in the 5G Era S.K. Sharma, S. Chatzinotas, and D. Arapoglou
Volume 80 Transceiver and System Design for Digital Communications 5th Edition Scott R.
Bullock
Volume 905 ISDN Applications in Education and Training R. Mason and P.D. Bacsich
Network Design, Modeling,
and Performance
Evaluation
Quoc-Tuan Vien
The Institution of Engineering and Technology

Published by The Institution of Engineering and Technology, London, United Kingdom
The Institution of Engineering and Technology is registered as a Charity in England &

Wales (no. 211014) and Scotland (no. SC038698).
© The Institution of Engineering and Technology 2019
First published 2018
This publication is copyright under the Berne Convention and the Universal Copyright
Convention. All rights reserved. Apart from any fair dealing for the purposes of research
or private study, or criticism or review, as permitted under the Copyright, Designs and
Patents Act 1988, this publication may be reproduced, stored or transmitted, in any
form or by any means, only with the prior permission in writing of the publishers, or in
the case of reprographic reproduction in accordance with the terms of licences issued
by the Copyright Licensing Agency. Enquiries concerning reproduction outside those
terms should be sent to the publisher at the undermentioned address:
The Institution of Engineering and Technology

Michael Faraday House
Six Hills Way, Stevenage
Herts, SG1 2AY, United Kingdom
www.theiet.org
While the author and publisher believe that the information and guidance given in this
work are correct, all parties must rely upon their own skill and judgement when making
use of them. Neither the author nor publisher assumes any liability to anyone for any
loss or damage caused by any error or omission in the work, whether such an error or
omission is the result of negligence or any other cause. Any and all such liability
is disclaimed.
The moral rights of the author to be identified as author of this work have been
asserted by him in accordance with the Copyright, Designs and Patents Act 1988.
British Library Cataloguing in Publication Data

A catalogue record for this product is available from the British Library
ISBN 978-1-78561-336-4 (hardback)

ISBN 978-1-78561-337-1 (PDF)
Typeset in India by MPS Limited

Printed in the UK by CPI Group (UK) Ltd, Croydon
I dedicate this book to my parents, Dich-Tung Vien and Thi-Ngoc Ngo,
my wife, Tram Pham, and our lovely children, Helen and Harry.
This page intentionally left blank
Contents
List of figures xiii

List of tables xvii
Foreword xix
Preface xxi
Abbreviations xxv
1 Internetworking and network fundamentals 1

1.1 Fundamentals of networking 1
1.2 Local area networks 5
1.2.1 LAN devices 5
1.2.2 LAN topologies 7
1.2.3 LAN technologies 9
1.2.4 LAN access methods 12
1.2.5 LAN transmission methods 14
1.3 Wide area networks 16
1.3.1 WAN devices 16
1.3.2 WAN topologies 18
1.3.3 WAN technologies 20
1.4 Open systems interconnection reference model 28
1.4.1 OSI layers 29
1.4.2 Interaction between OSI layers in communications 30
1.4.3 Information exchange and information formats in OSI
layers 31
1.5 Internet Protocols 32
1.5.1 TCP/IP suite 33
1.5.2 Internet Protocol 34
1.5.3 Transmission Control Protocol 34
1.5.4 User datagram protocol 34
1.5.5 Other protocols 34
1.6 Internetworks 35
1.6.1 Internetwork hierarchy 36
1.6.2 Internetwork addressing 36
1.6.3 Standards organizations 38
1.6.4 Internetworking devices 38
1.7 Related works 41
1.8 Review questions 41
1.9 Problems 42
viii Network design, modeling, and performance evaluation
2 Routing in computer networks 45

2.1 Fundamentals of routing 45
2.1.1 Path determination 46
2.1.2 Routing algorithms 46
2.2 Static routing protocols 48
2.3 Dynamic routing protocols 48
2.4 Interior gateway protocols 49
2.4.1 Distance-vector routing protocols 49
2.4.2 Link-state routing protocols 52
2.4.3 Enhanced Interior Gateway Routing Protocol 56
2.5 Border Gateway Protocols 56
2.8 Problems 58
3 Internet Protocol addressing 61

3.1 An overview of IP address 61
3.1.1 Addressing types 63
3.1.2 Addressing mechanisms 63
3.2 IP version 4 (IPv4) 63
3.2.1 Private IP addresses 64
3.2.2 Classful IP addresses 64
3.3 Subnetting 66
3.4 Variable-length subnetting 69
3.5 Supernetting 70
3.6 IP version 6 71
3.9 Problems 74
4 Network analysis, architecture, and design 79

4.1 An overview of network analysis, architecture, and design 79
4.2 Network analysis process 81
4.3 Network architecture process 81
4.4 Network design process 82
4.5 Network hierarchical design 83
4.6 Network hierarchical layers 84
4.6.1 Core layer 84
4.6.2 Distribution layer 84
4.6.3 Access layer 85
4.7 Network design approaches and evolution 86
4.8 Network management and security 87
4.8.1 Network design goals 88
4.8.2 Technology evaluation criteria 89
4.8.3 Network-management design 89
Contents ix

4.11 Problems 92
5 Network-requirement analysis 93
5.1 Network services and requirements 93
5.2 Service characteristics 94
5.3 Requirement-analysis process 98
5.4 User-requirement analysis 99
5.5 Application-requirement analysis 100
5.6 Host-requirement analysis 101
5.7 Network-requirement analysis 102
5.8 Requirement-analysis model 103
5.9 Reliability, maintainability, and availability analyses 104
5.12 Problems 111
6 Network flow analysis 115

6.1 Flows 115
6.2 Data sources and sinks 117
6.3 Flow models 117
6.3.1 Peer-to-peer model 118
6.3.2 Client–server model 119
6.3.3 Cooperative computing model 119
6.3.4 Distributed computing model 120
6.4 Flow boundaries 121
6.5 Flow distributions 122
6.6 Flow specification 123
6.7 Flow analysis model 125
6.10 Problems 128
7 Network performance evaluation 129

7.1 Benchmarking: merits and demerits 129
7.2 Simulation: merits and demerits 130
7.3 Analytical modeling: merits and demerits 131
7.4 Systems and system environment 132
7.7 Problems 135
8 Network simulation models 137

8.1 Simulation for emulating and evaluating a system 137
8.2 Simulation models of a system 139
x Network design, modeling, and performance evaluation
8.3 Steps/phases in a simulation 140

8.4 An example of simulation model 140
8.7 Problems 143
9 Statistical models in network simulation 145

9.1 Statistical models in queuing systems 145
9.2 Poisson point process 147
9.2.1 Stationary Poisson point process 148
9.2.2 Nonstationary Poisson point process 150
9.2.3 Operations of Poisson point process 151
9.3 Input modeling 152
9.3.1 Data collection 152
9.3.2 Probability distribution identification 153
9.3.3 Parameter estimation 154
9.3.4 Distribution fitness test 155
9.6 Problems 156
10 Probabilities in performance analysis 159

10.1 Basic concepts and terminology of probability 159
10.2 Axioms and properties of probability 161
10.3 Conditional probability 164
10.4 Independence of events 168
10.7 Problems 171
11 Random variables in network modeling and simulation 175

11.1 Introduction of random variables 175
11.2 Discrete random variables 176
11.2.1 Probability mass function 177
11.2.2 Cumulative distribution function 178
11.2.3 Expected value 180
11.2.4 Variance and standard deviation 182
11.2.5 Moments 185
11.3 Continuous random variables 186
11.3.1 Cumulative distribution function 187
11.3.2 Probability density function 188
11.3.3 Expected value 190
11.3.4 Variance and standard deviation 190
11.3.5 Moments 191
Contents xi

11.6 Problems 193
12 Discrete random variables and their applications 197

12.1 Review of discrete random variables 197
12.2 Bernoulli random variable 197
12.3 Binomial random variable 200
12.4 Geometric random variable 202
12.5 Pascal random variable 205
12.6 Poisson random variable 208
12.7 Discrete uniform random variable 211
12.10 Problems 215
13 Continuous random variables and their applications 217

13.1 Review of continuous random variables 217
13.2 Continuous uniform random variable 217
13.3 Exponential random variable 222
13.4 Erlang random variable 226
13.5 Normal random variable 230
13.6 Lognormal random variable 235
13.9 Problems 239
14 Random variable generation in network simulation 241

14.1 Inverse transform technique 241
14.1.1 Bernoulli random variable generation 242
14.1.2 Geometric random variable generation 243
14.1.3 Discrete uniform random variable generation 245
14.1.4 Continuous uniform random variable generation 246
14.1.5 Exponential random variable generation 247
14.2 Transformation techniques 248
14.2.1 Binomial random variable generation 248
14.2.2 Pascal random variable generation 249
14.2.3 Erlang random variable generation 250
14.2.4 Lognormal random variable generation 250
14.5 Problems 253
15 Queuing theory for network modeling and performance evaluation 255

15.1 An introduction to queuing theory 255
15.2 Queuing systems in computer and communication networks 256
xii Network design, modeling, and performance evaluation
15.3 Poisson point process for queuing systems 257

15.4 Performance measures in queuing systems 258
15.4.1 Interarrival and service time in Poisson point process 259
15.4.2 Waiting and response time 259
15.4.3 The number of packets in the queue and in the system 260
15.4.4 Traffic intensity and server utilization 260
15.4.5 Little’s law for queuing systems 261
15.5 Kendall’s notation for queuing systems 262
15.8 Problems 264
16 Single-server queues—network behaviors and analysis 267

16.1 M /M /1 Queue 267
16.2 M /M /1/K Queue 271
16.3 M/M/1/K/K Queue 274
16.6 Problems 279
17 Multi-server queues—network behaviors and analysis 281

17.1 M /M /c Queue 281
17.2 M /M /c/K Queue 286
17.3 M /M /c/K/K Queue 291
17.6 Problems 297
References 299
Index 315
List of figures
Figure P.1 Dependence graph of chapters xxiii

Figure 1.1 Local area network 2
Figure 1.2 Wide area network 3
Figure 1.3 Campus area network 3
Figure 1.4 Metropolitan area network 4
Figure 1.5 Home area network 4
Figure 1.6 Personal area network 5
Figure 1.7 Hub and its network symbol 6
Figure 1.8 Switch and its network symbol 6
Figure 1.9 Router and its network symbol 7
Figure 1.10 Bus topology 7
Figure 1.11 Star topology 8
Figure 1.12 Ring topology 8
Figure 1.13 Extended star topology 9
Figure 1.14 Unshielded twisted-pair (UTP) 10
Figure 1.15 Shielded twisted-pair (STP) 10
Figure 1.16 Optical fiber 10
Figure 1.17 Carrier sense multiple access with
collision detection (CSMA/CD) 12
Figure 1.18 Carrier sense multiple access with
collision avoidance (CSMA/CA) 13
Figure 1.19 Control token technique 14
Figure 1.20 Unicast transmission 15
Figure 1.21 Broadcast transmission 15
Figure 1.22 Multicast transmission 16
Figure 1.23 A WAN switch 17
Figure 1.24 Modems in WAN 17
Figure 1.25 A channel service unit/digital service unit (CSU/DSU) in WAN 18
Figure 1.26 Access server in WAN 18
Figure 1.27 A multiplexer in WAN 19
Figure 1.28 Full mesh topology 19
Figure 1.29 Partial mesh topology 20
Figure 1.30 Circuit switching 21
Figure 1.31 Packet switching 21
Figure 1.32 Virtual circuit switching 22
Figure 1.33 Integrated Services Digital Network (ISDN) devices 23
xiv Network design, modeling, and performance evaluation
Figure 1.34 X.25 devices 24

Figure 1.35 Frame relay devices 25
Figure 1.36 Asynchronous transfer mode (ATM) devices 26
Figure 1.37 Switched multimegabit data service (SMDS) devices 27
Figure 1.38 Digital subscriber line (DSL) 27
Figure 1.39 Open systems interconnection (OSI) layers 28
Figure 1.40 Upper and lower OSI layers 31
Figure 1.41 Protocol data unit (PDU) in OSI model 32
Figure 1.42 Transmission Control Protocol (TCP)/Internet Protocol (IP) 33
Figure 1.43 Internetwork hierarchy 36
Figure 1.44 MAC address 37
Figure 2.1 Interior gateway protocol (IGP) versus exterior
gateway protocol (EGP) 48
Figure 2.2 Routing loop in Example 2.1 52
Figure 2.3 Open Shortest Path First (OSPF) routing protocol 54
Figure 2.4 OSPF over backbone area 54
Figure 2.5 A graph in Example 2.2 where the value on the edge
represents weight or path cost 55
Figure 2.6 Border Gateway Protocols (BGPs) with Internal BGP (IBGP)
and External BGP (EBGP) 57
Figure 2.7 Figure for Problems 4 and 5 59
Figure 2.8 Figure for Problem 6 59
Figure 3.1 IP address with network mask in Example 3.1 62
Figure 3.2 IP address with subnet mask in Example 3.1 62
Figure 3.3 An IPv4 address 64
Figure 3.4 IP address classes 65
Figure 3.5 Subnetting 66
Figure 3.6 Variable-length subnet mask in Example 3.5 69
Figure 3.7 Supernetting in Example 3.8 71
Figure 3.8 IPv6 address in Example 3.9 72
Figure 4.1 Network analysis, architecture, and design process 80
Figure 4.2 Network hierarchical design with interconnectivity 83
Figure 4.3 Network hierarchical layers 84
Figure 4.4 One-layer design 85
Figure 4.5 Two-layer design 86
Figure 4.6 Network hierarchy over content-delivery network (CDN) 87
Figure 5.1 System components 94
Figure 5.2 Packet error probability (PER) versus the number of bits 97
Figure 5.3 Channel capacity versus signal-to-noise ratio (SNR) 98
Figure 5.4 Requirement-analysis process between two end nodes 99
Figure 5.5 Requirement-analysis model 103
Figure 5.6 Time between failure (TBF) and time to repair (TTR) of a system 105
Figure 5.7 Devices connected in series 106
List of figures xv
Figure 5.8 Devices connected in parallel 107

Figure 5.9 Figure for Problem 5. 111
Figure 6.1 Flow types 116
Figure 6.2 Peer-to-peer model 118
Figure 6.3 Client–server model 119
Figure 6.4 Cooperative computing model 120
Figure 6.5 Distributed computing model 121
Figure 6.6 Flow boundaries 122
Figure 6.7 Flow distribution 123
Figure 6.8 Flow analysis model 125
Figure 7.1 System components in a simulation model 132
Figure 7.2 Data flow between two end devices 133
Figure 7.3 Components in a banking system 134
Figure 7.4 Components in a single-server queuing system 134
Figure 8.1 Simulation phases 141
Figure 8.2 Simulation phases in a queuing system 142
Figure 9.1 Interarrival time in Poisson point process N (t) 149
Figure 9.2 Input modeling in network simulation 152
Figure 10.1 Event in an experiment 161
Figure 11.1 Random variable for representing a random event 175
Figure 11.2 pmf of random variable in Example 11.2 177
Figure 11.3 cdf of random variable in Example 11.2 179
Figure 11.4 cdf of a continuous uniform random variable in [1, 3] 187
Figure 11.5 pdf of a continuous uniform random variable in [1, 3] 188
Figure 12.1 pmf of X ∼ Ber(0.6) 199
Figure 12.2 pmf of X ∼ Bin(20, 0.6) 201
Figure 12.3 pmf of X ∼ Geo(0.6) 203
Figure 12.4 pmf of X ∼ Pas(4, 0.6) 206
Figure 12.5 pmf of X ∼ Poi(5) 209
Figure 12.6 pmf of X ∼ U (1, 5) 211
Figure 13.1 pdf of X ∼ U (1, 5) 219
Figure 13.2 cdf of X ∼ U (1, 5) 220
Figure 13.3 pdf of X ∼ Exp(5) 223
Figure 13.4 cdf of X ∼ Exp(5) 223
Figure 13.5 pdf of X ∼ Erl(3, 5) 227
Figure 13.6 cdf of X ∼ Erl(3, 5) 228
Figure 13.7 pdf of X ∼ N (1, 9) 232
Figure 13.8 cdf of X ∼ N (1, 9) 232
Figure 13.9 pdf of X ∼ LN (1, 0.25) 236
Figure 13.10 cdf of X ∼ LN (1, 0.25) 236
Figure 14.1 Samples of X ∼ Ber(0.6) 243
Figure 14.2 Samples of X ∼ Geo(0.6) 244
Figure 14.3 Samples of X ∼ U (1, 5), X ∈ Z 245
xvi Network design, modeling, and performance evaluation
Figure 14.4 Samples of X ∼ U (1, 5) 246

Figure 14.5 Samples of X ∼ Exp(5) 247
Figure 14.6 Samples of X ∼ Bin(20, 0.6) 248
Figure 14.7 Samples of X ∼ Pas(4, 0.6) 249
Figure 14.8 Samples of X ∼ Erl(3, 5) 250
Figure 14.9 Samples of Y ∼ N (1, 0.25) 251
Figure 14.10 Samples of Y ∼ LN (1, 0.25) 252
Figure 15.1 Components in a queuing system 257
Figure 15.2 Performance measures in a queuing system 262
Figure 16.1 M /M /1 queue 267
Figure 16.2 M /M /1/K queue 271
Figure 16.3 M /M /1/K/K queue 275
Figure 17.1 M /M /c Queue 282
Figure 17.2 M /M /c/K Queue 287
Figure 17.3 M /M /c/K/K Queue 292
List of tables
Table 2.1 Dijkstra’s algorithm for finding the shortest distance

in Example 2.2 56
Table 3.1 Private IP address range 64
Table 3.2 IP address classes 65
Table 3.3 Subnetting in Example 3.3 67
Table 3.4 Laboratories in Example 3.4 68
Table 3.5 Subnetting in Example 3.4 68
Table 3.6 Rooms in Example 3.6 70
Table 3.7 Variable-length subnetting in Example 3.6 70
Table 3.8 IPv6 packet format 73
Table 3.9 Subnetting of Class C address 74
Table 3.10 Subnetting of Class B address 75
Table 3.11 List of workgroups in Problem 12 76
Table 3.12 List of departments in Problem 14 77
Table 5.1 Table for Problem 5 112
Table 12.1 Formulas of discrete random variables 198
Table 13.1 Formulas of continuous random variables 218
Foreword
About the book
This book introduces a methodological approach to network design that enables read-
ers to evaluate a network situation and identify the most important aspects to be
monitored and analyzed. It shows how to go from the analysis of initial network
requirements to the network architecture design, modeling, simulation, and evalua-
tion. Practice exercises are given for selected chapters, and case studies take the reader
through the whole network-design process. It is ideal for practitioners and researchers
working in all aspects of network services.
About the author

Quoc-Tuan Vien is currently a senior lecturer in computing and communications
engineering within the Faculty of Science andTechnology, Middlesex University, Lon-
don, UK. Prior to joining Middlesex University, he worked as postdoctoral research
assistant with the School of Science and Technology, Nottingham Trent University,
Nottingham, UK. He received his PhD degree from Glasgow Caledonian University,
UK, in 2012.
Dr Vien is a senior member of the IEEE, a member of the IET, and a fellow of
the HEA. He has authored/coauthored two books, five book chapters, and over 70
publications in major conference proceedings and ISI journals in which he is a leading
author of 43 papers. His current research interests include physical layer security, net-
work coding, non-orthogonal multiple access, energy harvesting, spectrum sensing,
device-to-device communications, relay networks, cognitive radio networks, hetero-
geneous networks, wireless network-on-chip, public safety networks, and cloud radio
access networks.
He was a recipient of the Best Paper Award at the IEEE/IFIP 14th International
Conference on Embedded and Ubiquitous Computing in 2016. He has been an edi-
tor of International Journal of Digital Multimedia Broadcasting, a guest editor of
the EAI Endorsed Transactions on Industrial Networks and Intelligent Systems, a
program cochair for the EAI International Conference on Industrial Networks and
Intelligent Systems (INISCOM 2018), a technical symposium cochair for the Interna-
tional Conference on Recent Advances in Signal Processing, Telecommunications and
Computing (SigTelCom 2017, 2018), and aTPC member of over 100 conferences such
as the IEEE PIMRC, ICNC, VTC, ISWCS, ATC, etc. He has also served as a session
chair at IEEE flagship conferences including the IEEE WCNC, VTC, and ISWCS. He
was honored as an exemplary reviewer of the IEEE Communications Letters in 2017.
Preface
About the book

Over the last five decades, internetworking has gained significant interest in con-
necting people with people and machines with machines through computer networks.
Internetworking provides sustainable links between various types of network topology
connecting from a small local area network to a wide area network. As a remarkable
example of internetworking, the Internet has played an important role in our daily life
with apparent benefits. By surfing the web, we are facing the outside world with up-
to-date news, connecting with old friends, and knowing new friends that geographical
distance does not allow us to meet face-to-face. Imagine you are planning to travel
and what happens without the Internet? In fact, with the Internet, your travel would be
much simpler and enjoyable when everything can be organized according to a routine,
such as booking airplane tickets, reserving hotel rooms, tracking maps, checking the
weather forecast, researching the culture of the new area, etc. All of these conve-
niences depend on the network-architecture design and modeling, which require not
only a background knowledge but also extra effort and investment on the learning of
various models and techniques to analyze their effectiveness in practical situations.
Intended audience
This book is intentionally designed for undergraduate and graduate programs in net-
working, communications, and computer engineering, as well as communications,
IT, and networking professionals, researchers and scientists involved in the planning,
design, development, testing, and operation of network services. This book is also
suitable for self-study to gain a basic knowledge of computer-network design to be
able to develop a simple network model for further analysis at advanced level.
Developed as a textbook, this book can benefit lecturers and students in net-
working, communications, and computer engineering. The chapters are particularly
designed as a series of independent modules that can be combined in a number of
ways for designing the courses.
Plan of the book

This book will help you to understand how to evaluate a network situation and to
identify the most important network aspects that need to be monitored and analyzed.
This will be done through the introduction of a methodological approach to network
design. This book will also introduce the concepts of network modeling, analysis and
xxii Network design, modeling, and performance evaluation
simulation, and examine the techniques facilitating such work. It will provide you with
the appreciation of the design and development simulation software as appropriate
models to evaluate the pure performance and availability, as well as performability of
computer networks.
This book covers the following topics:
● Internetworking and internetworking units
● Systematic approach to network design:
– Requirements and flow analysis
– Logical design
– Addressing and routing
– Internetworking strategies: switching, routing, hierarchy, and redundancy
● Probability and statistics used in modeling networks:
– Probabilities
– Random variables and their families
– Random-variate generation
– Application of statistical models in network simulation
● Input modeling techniques and output data analysis
● Queueing theory and analytical modeling of nodes and networks:
– Queueing models and their variations
– Analysis of queueing models and their applications in network design
● Performance evaluation of computer networks
– Network simulation
– Modeling for analytical solutions and simulation
– Modeling and simulation of single-and multi-server queueing systems
– Modeling of complex systems
– Building simulation models and analytical models
In this book, I will provide detailed steps from the analysis of initial network
requirements to the network architecture design, modeling, simulation, and evalua-
tion. In particular, there are more focuses on the statistical and queueing models to
show that they are particularly useful for the network design and analysis. This book
is directed at readers with little or no background in networking. For the readers with
great interest, related works are provided in a separate section of every chapter for
further study. Also, this book contains a number of review questions and problems at
the end of each chapter for the readers to practice.
Usage of the book
This book is specifically designed for undergraduate and graduate programs in net-
working, communications, and computer engineering, as well as communications,
IT, and networking professionals, researchers and scientists involved in the planning,
design, development, testing, and operation of network services.
For best use of the book, the students of undergraduate programs in network-
ing, communications, and computer engineering are suggested to follow the whole
structure of the book. The students of postgraduate programs who are supposed to
Preface xxiii
have good background knowledge in computer networks and used to work on network
design/development can have a quick review of Chapters 1–3.
The book consists of several homework problems that are organized to assist both
instructors and students with reference to the relevant sections.
Dependence graph
The dependence between chapters in the book is summarized in Figure P.1 where
some chapters can be skipped depending on the background of the students and also
on the preferences of the instructor.
Chapter 1
Background
Chapter 2 of
networking
Chapter 3 Network
design
Chapter 4 Chapter 5
Chapter 6
Chapter 7
Chapter 8 Modeling
Chapter 9
Chapter 10
Chapter 11
Chapter 12
Performance
evaluation
Chapter 13 Chapter 15
Chapter 14 Chapter 16
Chapter 17
Figure P.1 Dependence graph of chapters

Abbreviations
ABR area border router

ADSL asymmetric digital subscriber line
ANSI American National Standards Institute
AP access point
ARP address resolution protocol
AS autonomous system
ASCII American Standard Code for Information Interchange
ATM asynchronous transfer mode
BER bit error rate
BGP Border Gateway Protocol
bps bits per second
BRI Basic Rate Interface
CAN campus-area network
CAPEX CAPital EXpenditures
CCITT Consultative Committee for International
Telephony and Telegraphy
cdf cumulative distribution function
CDN content delivery network
CIDR classless inter-domain routing
CIR committed information rate
CoS classes of service
CPE customer premises equipment
CSMA/CA carrier sense multiple access with collision avoidance
CSMA/CD carrier sense multiple access with collision detection
CSU channel service unit
CTS clear-to-send
DARPA Defense Advanced Research Projects Agency
DCE data circuit-terminating equipment
DDoS distributed denial-of-service
DHCP Dynamic Host Configuration Protocol
DNS domain name system
DSL digital subscriber line
DSU digital service unit
DTE data terminal equipment
DUAL diffusing update algorithm
DVRP distance-vector routing protocol
xxvi Network design, modeling, and performance evaluation
EBCDIC Extended Binary Coded Decimal Interchange Code

EBGP External Border Gateway Protocol
ECMA European Computer Manufacturing Association
EGP Exterior Gateway Protocol
EIGRP Enhanced Interior Gateway Routing Protocol
ES end system
FCFS first-come first-served
FIFO first-in first-out
FIRO first-in random-out
FTP File Transfer Protocol
Gbps gigabits per second
GHz gigahertz
GIF Graphics Interchange Format
GIGO garbage-in–garbage-out
HAN home-area network
HDSL high bit-rate digital subscriber line
HTTP HyperText Transfer Protocol
Hz hertz
IAB Internet Activities Board
IANA Internet Assigned Numbers Authority
IBGP Internal Border Gateway Protocol
ICMP Internet Control Message Protocol
IEEE Institute of Electrical and Electronics Engineers
IETF Internet Engineering Task Force
IGP interior gateway protocol
IGRP Interior Gateway Routing Protocol
IP Internet Protocol
IPv4 Internet Protocol version 4
IPv6 Internet Protocol version 6
IS intermediate system
IS-IS intermediate system to intermediate system
ISDN Integrated Services Digital Network
ISO International Organisation for Standardization
ITU International Telecommunication Union
ITU-T ITU-Telecommunication Standardization Sector
IWU internetworking unit
JPEG Joint Photographic Experts Group
Kbps kilobits per second
LAN local-area network
LIFO last-in first-out
LLC Logical Link Control
LSRP link-state routing protocol
MAC media access control
MAN metropolitan-area network
Abbreviations xxvii
Mbps megabits per second

MIP management information base
MPEG Motion Picture Experts Group
MTBF mean time between failures
MTTR mean time to repair
MTU maximum transmission unit
MUX multiplexer
NAT network address translator
NIC network interface card
NPPP Nonstationary Poisson Point Process
NT network termination
OPEX operating expenses
OSI open systems interconnection
OSPF Open Shortest Path First
P2P peer-to-peer
PAN personal-area network
PC personal computer
pdf probability density function
PDN public data network
PDU protocol data unit
PER packet error rate
pmf probability mass function
PPP Poisson point process
PRI Primary Rate Interface
PSE packet-switching exchange
PSN packet-switched network
PSTN public switched telephone network
PVC permanent virtual circuit
Q–Q quantile–quantile
QoS quality of service
RARP Reverse Address Resolution Protocol
RFC Request For Comments
RIP Routing Information Protocol
RMA reliability, maintainability and availability
RTS request-to-send
SDSL symmetric digital subscriber line
SDU service data unit
SIP SMDS Interface Protocol
SMDS switched multimegabit data service
SMTP Simple Mail Transfer Protocol
SNI Subscriber Network Interface
SNMP Simple Network-Management Protocol
SNR signal-to-noise ratio
SPPP stationary Poisson point process
xxviii Network design, modeling, and performance evaluation
STP shielded twisted pair

SVC switched virtual circuit
TA terminal adapter
TCP Transmission Control Protocol
TE terminal equipment
TFTP Trivial File Transfer Protocol
TIFF Tagged Image File Format
ToS types of service
UDP User Datagram Protocol
UTP unshielded twisted pair
VC virtual circuit
VDSL very-high-data-rate digital subscriber line
VLSM Variable-Length Subnet Mask
WAN wide-area network
WLAN wireless local-area network
WWW World Wide Web
Chapter 1
Internetworking and network fundamentals
It has been decades since the first computer was produced to do a very basic cal-
culation. When first developed, the computer was simply a standalone machine to
help people in either solving problems that require high-complexity computational
algorithms or storing data or providing amusement with a variety of entertainment
and was therefore named as a personal computer (PC). Over the time, such PC has
been keeping changing to meet the users’ non-stopping requirements in daily life.
Along with a number of social activities of high interactive demands between users
of distant areas, a two-way flow of information between the PC and the user is shown
to be insufficient and hence necessitate the networking concept to link all the PCs
in a proper manner that allows them to understand each other even though they have
different configurations and may be located at different continents.
Aiming at showing an overview picture of computer networks, in this chapter,
I will first outline the fundamentals of networking in Section 1.1 with different net-
work models. Starting from small-size networks, Section 1.2 will discuss local area
network (LAN) for limited geographic area with variant topologies, technologies and
access mechanisms. Section 1.3 will then consider wide area network (WAN) for
a large geographic area, where I will highlight some well-known WAN devices and
topologies along with relevant technologies over network media. In order to model the
communications between different devices in different networks, a reference model
or framework is required. Section 1.4 will provide a brief explanation of seven layers
in open systems interconnection (OSI) framework. As the most widely used pro-
tocol over the Internet, Transmission Control Protocol/Internet Protocol (TCP/IP)
suite with a variety of communication protocols will be presented in Section 1.5.
A brief account of internetworks and internetworking units will be explained in Sec-
tion 1.6 where I will show the functionalities of intermediary devices for computer
networking, including repeaters/hubs, bridges/switches, and routers/gateways.
1.1 Fundamentals of networking

In the early days of computer networks, only data was shared between connected
PCs of the same types. A simplest network is when we connect two PCs via either
wire or wireless to share data. With the advance in communication technology, more
computer systems can be linked together to make a complex computer network. The
2 Network design, modeling, and performance evaluation
evolved technology also allowed sharing graphic, voice and even video streams over
various types of computer systems. In the 1990s, the Internet was emerged as a
novel concept to form a global community as a global communication network where
people at any location or time zone can simply surf the World Wide Web (WWW) to
exchange the information. Since then, the Internet has become a part of our daily life.
However, behind such high-technology communication network, there are a number
of interconnections between electronic devices and computer systems that we are not
aware of their presence.
Depending on the size, service, and functionality of a network, there are many
types of computer networks as follows:
● LAN (cf. Figure 1.1): connects the devices which are geographically close to
each other. As an extension of wired LAN, wireless LAN (WLAN) enables the
over-the-air connection of wireless enabled devices.
● WAN (cf. Figure 1.2): consists of two or more LANs connecting the devices
which are geographically far apart.
● Campus area network (CAN) (cf. Figure 1.3): connects LANs within either a
school campus or an enterprise campus.
● Metropolitan area network (MAN) (cf. Figure 1.4): connects LANs within a town
or a city but has a smaller coverage area compared to WAN.
● Home area network (HAN) (cf. Figure 1.5): connects devices within an individual
home.
● Personal area network (PAN) (cf. Figure 1.6): is the smallest network connecting
devices around an individual person, which could be PCs, telephones, mobile
phones, tablets, personal digital assistants (PDAs), printers, etc.
Figure 1.1 Local area network

Internetworking and network fundamentals 3
Figure 1.2 Wide area network
University
Figure 1.3 Campus area network

Figure 1.4 Metropolitan area network
Figure 1.5 Home area network
As listed above, the size of the geographical area is exploited to name distinct
groups of computer networks. In the scope of this book, I will scrutinize the LANs
and WANs in most parts, while the CANs, MANs, PANs, and HANs can be referred
to as special models with specific group of users and geographic boundaries.
Figure 1.6 Personal area network
1.2 Local area networks

A LAN is first defined as a group of PCs connected within a limited geographic area,
e.g., in one floor of a building (cf. Figure 1.1). With the development of technology,
the LAN can be referred to a larger group of devices in different floors of the building.
Over the LANs, given a short communication distance, it is expected that the
devices can communicate with each other at a high data rate. In other words, a high
bandwidth (cf. Definition 1.1) can be allocated in the LANs. The LANs also allow
multiple access with full-time connectivity of different devices which can be locally
managed at the access point (AP) in a centralized manner or distributed manner
without the AP.
Definition 1.1. Bandwidth, in computing, is defined as the maximum data rate or
information capacity having units of bits per second (bps).1
1.2.1 LAN devices

In order to connect devices within a LAN, appropriate devices should be selected.
There are three well-known LAN devices including hubs, switches, and routers. In
this subsection, a brief introduction of these devices is provided for an overview of
their functionality, while I will discuss about them in detail in Section 1.6.4.
1.2.1.1 Hubs
Hubs are the basic intermediate network devices to link end network devices to a
shared medium in a small LAN with low throughput requirements (cf. Definition 1.2).
1
Note that the bandwidth is defined differently in digital communications, which is used to represent the
operation frequency range measured in hertz (Hz). In the rest of this book, bandwidth and capacity are
used interchangeably, unless otherwise indicated.
Network symbol
Figure 1.7 Hub and its network symbol
Network symbol
Figure 1.8 Switch and its network symbol
A simplest form of the hubs is a repeater with only two physical incoming and outgoing
ports, which used to extend the network coverage. The hubs are generally regarded
as multi-port repeaters to regenerate the received signals and send them over all ports
(cf. Figure 1.7).
Definition 1.2. Throughput is defined as the actual data transfer rate in bps.
1.2.1.2 Switches
Switches are normally used in a large network to connect devices to a LAN by regener-
ating the received data frame and sending it to a certain destination port (cf. Figure 1.8).
A predecessor of switches is a bridge that simply connects a LAN to another LAN
via two physical ports. The switches typically have multiple ports and thus can be
regarded as multi-port bridges.
1.2.1.3 Routers
Routers are intermediary network devices used to connect LANs and WANs (cf.
Figure 1.9). The routers receive and forward data packets based on addresses. Instead
of using hardware to forward the packets as in switches, the routers use software
to manage the packet forwarding with different protocols and also support different
network technologies.
Network symbol
Figure 1.9 Router and its network symbol
Figure 1.10 Bus topology
1.2.2 LAN topologies

In a LAN, devices are connected via physical links following different geometric
shapes. Apart from the simplest point-to-point connection, electronic devices and
LAN intermediary devices can be connected via various physical topologies, which
can be listed in the following sections.
1.2.2.1 Bus
Bus topology has been deployed for connecting PCs since the early implementation of
historic Ethernet where all devices are connected to a single central cable, namely, the
bus, as the backbone of the network and they use a shared medium (cf. Figure 1.10).
1.2.2.2 Star
In a star topology, devices are connected to a central point of the network, e.g., hubs,
over the same shared medium as in bus topology (cf. Figure 1.11).
1.2.2.3 Ring
In a ring topology, devices are linked together in a closed loop with no central point.
Each device sequentially receives and passes the data frames to the next device until
reaching the destination (cf. Figure 1.12).
Figure 1.11 Star topology
Figure 1.12 Ring topology

Figure 1.13 Extended star topology
1.2.2.4 Extended star

Extended star topology is used to enlarge the coverage area of the star topology by
using additional repeaters/hubs (cf. Figure 1.13).
1.2.3 LAN technologies

In order to assist the communication between workstations, PCs, printers, servers,
etc., a combination of both software and hardware is required with technologies to
maintain the traffic in the shared network. Basically, there are two LAN technologies
including Ethernet and token ring.
As a family of LAN products, Ethernet is the most widely used technology
defined by the Institute of Electrical and Electronics Engineers (IEEE). The Ethernet
has sustained for over 35 years since the first Ethernet standard with the first products
were developed in 1980. The Ethernet has been a fast and reliable network solution
for bus and star topology (cf. Sections 1.2.2.1 and 1.2.2.2) with evolved technologies
that can provide data speed in the range of 10 Mbps to 10 Gbp. There are different
types of the Ethernet standard depending on cabling and transmission media, which
can be over either copper cable with unshielded twisted-pair (UTP) (cf. Figure 1.14)
or shielded twisted-pair (STP) (cf. Figure 1.15), or optical fiber (cf. Figure 1.16) or
wireless. The current Ethernet media options include two general types of copper
cable: UTP and STP, plus several types of optical fiber cable.
In the following, I will describe some well-known Ethernet covered by the IEEE
802.3 and IEEE 802.11 standards, followed by an overview of token ring technology
and IEEE 802.5 standard for ring topology.
Figure 1.14 Unshielded twisted-pair (UTP)
Figure 1.15 Shielded twisted-pair (STP)
Figure 1.16 Optical fiber
1.2.3.1 10BaseT Ethernet

10BaseT or 10-Mbps Ethernet is a standard IEEE 802.3 Ethernet for connecting
devices in a LAN. It uses two pairs of UTP Category-3 cables to carry 10 Mbps
half-duplex or full-duplex communications in a star topology over a distance of up
to 100 m in length. In a half-duplex system, the communication between two devices

is realized separately but not simultaneously, whereas the devices in a full-duplex
system can communicate with each other simultaneously at the same time.
1.2.3.2 100BaseTX and 100BaseFX fast Ethernet
Known as fast Ethernet, 100BaseTX uses two pairs of UTP Category-5 or Type-1
STP cables, while 100BaseFX employs fiber-optic cables to transmit data at a rate
of 100 Mbps for either half-duplex or full-duplex communications. The fast Ethernet
uses the star topology and also supports the 10BaseT standard Ethernet networks.
1.2.3.3 1000BaseT and 1000BaseX gigabit Ethernet
1000BaseT Gigabit Ethernet provides 1 Gbps full-duplex transmission over a distance
of up to 100 m by employing four pairs of UTP Category-5 cable. Instead of using UTP
cables, another Gigabit Ethernet standard, namely, 1000BaseX, was also proposed to
use fiber-optic cables exploiting the advantages of optical fiber in data transmission
for high reliability and long distance as well as its extendibility to a much higher data
rate in later version of Ethernet standards.
1.2.3.4 10GBaseSR and 10GBaseLR 10Gbps Ethernet
One of the latest developments of Ethernet standards to meet the high-bandwidth
requirements is the usage of fiber-optic cables to achieve a data rate of up to
10 Gbps, which is also known as 10 Gigabit Ethernet or 10GbE. Specifically, there
are 10GBaseSR and 10GBaseLR standards, where the 10GBaseSR employs multi-
mode fiber-optic cables for a short-range transmission of up to 300 m in a LAN or a
MAN, whereas 10GBaseLR makes use of single-mode fiber-optic cables to provide
a transmission range of up to 40 km and thus can be deployed in different LANs,
MANs, and even WANs.
1.2.3.5 Wireless Ethernet
Commonly referred to as Wi-Fi or WiFi, the IEEE 802.11 is a WLAN technology to
establish a wireless connection between devices for LAN extension, cross-building
access, nomadic access and ad hoc networks. In order to enable and maintain the
wireless connectivity function, the WLAN requires a wireless AP and a wireless
network interface card (NIC) adapter to provide wireless communication capability
to each device. The WLANs have emerged with various IEEE 802.11 standards in
use with different specifications to operate at different frequency bands, e.g., 2.4 GHz
(802.11b/g/n), 3.65 GHz (802.11y), 4.9 GHz (802.11j), 5 GHz (802.11a/n/h/j/ax/ac),
5.9 GHz (802.11p), and 60 GHz (802.11ad/aj/ay). Another technology for short-range
wireless communication is Bluetooth which is normally used in PANs (cf. Section 1.1).
The Bluetooth follows the IEEE 802.15 standards.
1.2.3.6 Token ring
Originally developed by IBM in the 1970s, Token Ring network has been known as
IBM’s LAN technology. After IBM, the IEEE also proposed a related IEEE 802.5
standard which is regarded as a shadow of the token ring. The IEEE 802.5 and the
token ring are interchangeable and compatible with each other. Both standards are
employed in token-passing networks where a small frame, namely, token, is moved

around the network and only the device keeping the token has the right to transmit.
1.2.4 LAN access methods

Over the shared media of a LAN with a number of devices, contention is likely to
take place when these devices all want to send data at the same time. It is vital to have
methods to manage the media access of the devices in the network. In order to answer
the question how the devices in a LAN access the network and share the medium, in
this subsection, I will discuss three access control methods, including carrier sense
multiple access with collision detection (CSMA/CD), carrier sense multiple access
with collision avoidance (CSMA/CA), and control token.
1.2.4.1 Carrier sense multiple access with collision detection

CSMA/CD is currently used in IEEE 802.3 Ethernet to deal with the contention when
transferring data in a LAN. The CSMA/CD detects collisions in a bus topology (cf.
Section 1.2.2.1) of the wired networks. When a device wants to transmit information, it
first listens to see if any other device(s) is currently transmitting data over the network
(cf. Figure 1.17). If the medium is free with no data transmission in progress, the device
Carrier sense
(CS)
Multiple access
(MA)
Collision detection
(CD)
Figure 1.17 Carrier sense multiple access with collision detection (CSMA/CD)
can transmit its data. Otherwise, the device has to wait until the occupied transmission
finishes prior to transmitting its data. However, there is a likely scenario that collision
may happen when two devices in the network transmit data simultaneously. In that
case, both devices back off, and each device will wait a random time period before
retransmitting its data. Such collision indeed occurs quite often in a network with
many devices and thus causes performance degradation of the Ethernet in a busy
network.
1.2.4.2 Carrier sense multiple access with collision avoidance

CSMA/CA is especially developed for the wireless medium in IEEE 802.11 WLAN
(cf. Section 1.2.3.5). The CSMA/CA is basically similar to the CSMA/CD for wired
LAN in carrier sensing, but devices in the WLAN have to first request for the right to
send to avoid collisions. In order to send data, the sender transmits a request-to-send
(RTS) frame to the receiver indicating the time required for data transmission and the
receiver sends a clear-to-send (CTS) frame to let the sender know that the channel
is reserved for the data transmission (cf. Figure 1.18). The RTS and CTS frames are
therefore of great importance in the WLAN with CSMA/CA to avoid the collisions.
1.2.4.3 Control token

Control token is a media access technique used mostly in a ring topology (cf. Section
1.2.2.3) with token ring technology and the IEEE 802.5 standard (cf. Section 1.2.3.6).
Each device sends its data only if it has a token which is a small frame used to manage
the token passing in the network (cf. Figure 1.19). The token is moved around the
Carrier sense
(CS)
RTS
Multiple access CTS

(MA)
Collision avoidance
(CA)
Data
Figure 1.18 Carrier sense multiple access with collision avoidance (CSMA/CA)
Token
Figure 1.19 Control token technique
network. If a device receives the token, it is granted the right to transmit data. However,
if the device holding the token does not want to send any data then it will pass the token
to the next device. Such token passing helps prevent the collisions in the network.
The data frame circulates the ring until it reaches the intended receiver. The receiver
copies the data frame for further processing and forward it to continue circulating
the ring until it reaches the sender for verifying if has been received by the receiver.
Although the collisions are unlikely to occur with control token, the delay caused
by circulating the token around the network is substantial in a large network with a
number of devices.
1.2.5 LAN transmission methods

Data transmission within a LAN can be classified into three essential types, including
unicast, multicast, and broadcast, which can be defined as follows.
1.2.5.1 Unicast transmission

Unicast transmission is used to express the data communication between one device
and another device (cf. Figure 1.20). There are only one sender and one receiver. The
Figure 1.20 Unicast transmission
Figure 1.21 Broadcast transmission
sender sends the data packet across the network to a specified receiver based on the
destination address. The unicast transmission is still of greatest importance in LAN.
1.2.5.2 Broadcast transmission

Broadcast transmission represents the data communication where a device sends data
to all other devices in the network (cf. Figure 1.21). There is only one sender, while all
other devices are receivers interested in receiving data. In the broadcast transmission,
the data packet is sent with a broadcast address and its copies will be received by all
the receivers.
1.2.5.3 Multicast transmission

Multicast transmission indicates the data communication where one or more devices
send data to a specific set of other devices (cf. Figure 1.22). There are one set of senders
and another set of receivers. The destinations of the data packet are determined with
Not interested
Not registered
Figure 1.22 Multicast transmission
a multicast address and a copy of the packet will be sent to each receiver in the set of
the multicast destination address.
1.3 Wide area networks
A WAN can be simply understood as a large data communication network connecting

two or more LANs to cover a broad geographical area (cf. Figure 1.2). The WAN
can be spread across multiple cities and countries and even in different continents.
Devices in the WAN are typically connected via common carrier circuits of service
providers, i.e., telephone/mobile phone companies, with serial links allowing them
to connect to public networks, i.e., the Internet.
1.3.1 WAN devices

There are different types of devices using in a WAN depending on the deployed
technologies. In the following, I will describe some common WAN devices, including
WAN switches, access servers, modems, channel service unit (CSU)/digital service
unit (DSU), and multiplexers (MUXs). Other specific devices with respect to each
WAN technology will be discussed in Section 1.3.3.
WAN
switch
Figure 1.23 A WAN switch
Modem Modem
Figure 1.24 Modems in WAN
1.3.1.1 WAN switches

Similar to a LAN switch (cf. Section 1.2.1.2), a WAN switch is a multi-port intermedi-
ary device having the same function of device connection (cf. Figure 1.23). However,
instead of connecting devices in the LAN, the WAN switches are used in carrier net-
works to direct traffic of different WAN technologies, e.g., frame relay, X.25, and
switched multimegabit data service (SMDS) which will be discussed in Section 1.3.3.
1.3.1.2 Modems
A modem is basically a device that has the function of both modulator and demodulator
to convey analog signals over the medium, e.g., telephone lines. A modem at the sender
converts the digital data to analog signals for transmission and another modem at the
receiver will do the reverse job by changing the received analog signals into the
original digital data (cf. Figure 1.24). Depending on the modulation/demodulation,
the modems may have different data rates.
1.3.1.3 Channel service unit/digital service unit

A CSU/DSU is a device used to convert a digital data frame from a router of a LAN
into a frame suitable for digital circuits used in a WAN and vice versa from a WAN
to a LAN (cf. Figure 1.25). The CSU/DSU processes the signals transmitted to or
received from the WAN, manages line control, and also provides signal timing for the
communication between devices in the LAN and the WAN.
WAN
switch
CSU/DSU
Figure 1.25 A channel service unit/digital service unit (CSU/DSU) in WAN
Modem
Modem Access
server
Modem
Figure 1.26 Access server in WAN
1.3.1.4 Access server

An access server is a center managing dial-up calls from the devices or hosts requesting
to login an Internet service or dial-out connections into a WAN, and vice versa with
dial-in connections (cf. Figure 1.26).
1.3.1.5 Multiplexers
A MUX is a device used to combine data channels from different WAN technologies
into a single data stream which can be then transmitted over a common line with an
integrated CSU/DSU (cf. Figure 1.27). The MUX also allows us to combine both
data and voice over the same line connecting to the WAN. The MUX can therefore
eliminate a number of individual connections for each channel saving significantly
operation and implementation costs.
1.3.2 WAN topologies

A WAN is used to connect multiple remote LANs of different sites using the inter-
mediary devices as described in Section 1.3.1. There are numerous approaches to
connect these WAN devices with different topologies depending on their locations.
Apart from the basic point-to-point, star, and ring topologies as in the LAN (cf. Sec-
tions 1.2.2.2 and 1.2.2.3), there are hybrid and mesh topologies. As the name says, a
MUX
Figure 1.27 A multiplexer in WAN
Figure 1.28 Full mesh topology
hybrid topology is simply a combination of two or more basic topologies, while, in

a mesh network, the devices are connected to each other over different links. In the
following, I will discuss two basic types of mesh topology including full mesh and
half mesh topologies.
1.3.2.1 Full mesh

In a full-mesh WAN (cf. Figure 1.28), each device has a connection to all other
devices in the network. With a number of redundant links, this topology provides the
best performance with the lowest possibility of failure as a single broken link does
not cause much affect on the data transmission. However, the full mesh connection
Figure 1.29 Partial mesh topology
can be applied only in a small network, whereas it is impractical in a large WAN with
a large number of devices.
1.3.2.2 Partial mesh

A partial mesh topology is deployed to partially connect devices with each other in a
WAN so that there are at least two devices connected via two or more links to other
devices in the network (cf. Figure 1.29). Such topology helps to reduce the cost of
the redundancy implementation in the network compared to a full mesh topology.
1.3.3 WAN technologies

In a WAN, to connect all devices around the world, simply using physical devices
and cables is insufficient, but that is a complicated process which requires numerous
technologies to sustain and maintain the communications. In this subsection, I will
first discuss two foundational technologies for switching, including circuit switching
and packet switching, along with connectionless and connection-oriented services
with virtual circuits (VCs). Then, I will provide an overview of various underlying
technologies and protocols associated with WANs, e.g., Integrated Services Digital
Network (ISDN), X.25, frame relay, asynchronous transfer mode (ATM), SMDS, and
digital subscriber line (DSL).
Figure 1.30 Circuit switching
3
2
Figure 1.31 Packet switching
1.3.3.1 Circuit switching

Circuit switching is a simple way to send data from a sender to a receiver by physically
connecting them when a data communication is required between them (cf. Fig-
ure 1.30). Such connection has to be established prior to transmitting data at the sender
and will last until the communication finishes. For this reason, the circuit-switched
communications are also referred to as connection-oriented communications. The
switched circuits operate as in a public switched telephone network for voice com-
munication. A typical instance of switched circuits can also be found in ISDN that
will be described in detail in Section 1.3.3.4.
1.3.3.2 Packet switching

Packet switching is a fundamental WAN technology for digital communications where
all devices transmit data in the form of packets over the shared medium with common
carrier resources (cf. Figure 1.31). Since the packets can be delivered in either con-
nectionless or connection-oriented manners, there are two types of packet switching,
including connectionless packet switching, also known as datagram packet switch-
ing, and connection-oriented packet switching, also known as VC packet switching.
This accordingly increases network efficiency and saves significantly the cost of
the customers in leasing the point-to-point lines from a carrier network provider. In
packet-switched networks (PSNs) with VCs, connections are virtually set up between
the WANs and the carrier network allowing the customers in different WANs share
the same carrier network which is referred to as a cloud. Some examples of the PSNs
include X.25, frame relay, ATM, and SMDS, which will be sequentially discussed in
Sections 1.3.3.5–1.3.3.8.
1.3.3.3 Virtual circuits

In PSNs, the carrier network provider set up a virtual connection between customer
sites in different WANs to deliver data packets between them. The virtual connection
is actually realized via a number of logical circuits within the shared carrier network
and is therefore named as VCs (cf. Figure 1.32). Depending on the frequency of usage
of the VC, following are the two types of VC:
● SwitchedVC (SVC): consists of circuit establishment, data transfer, and circuit ter-
mination. The SVCs are dynamically set up when required for data transmission,
so they are used for sporadic communication which occurs at irregular intervals or
only in a few places. Such dynamic process, although requires a high bandwidth
use and resource for establishment and termination, nevertheless saves the cost
of constantly setting up VC.
● Permanent VC (PVC): consists of only data transfer mode which is realized using
a permanently established VC. The PVCs are used in the situation when constant
data transmission is required between two remote customer sites. The PVCs help
save bandwidth use with the preset VCs, however result a considerably high cost
for constantly maintaining the availability of the VCs.
1.3.3.4 Integrated Services Digital Network

ISDN supports both digital telephone and data transport services through the usage
of existing telephone wires provided by telephone companies. Basically, in the ISDN,
2 1
3
Figure 1.32 Virtual circuit switching

when a router has voice, text, data, etc. to transmit to another router at a remote
site, an ISDN circuit is firstly initiated to set up a voice call to the remote network
with a telephone number. The data communication between the two devices will be
carried out when the connection between them is authentically established. As in
circuit switching, the connection will be terminated when the data communication is
complete.
Applications of ISDN involve providing additional telephone lines in homes for
tasks requiring integrated voice and data, high-speed file transfer, resource sharing,
and video-conferencing.
ISDN devices include the following (cf. Figure 1.33):
● Terminals: There are two types of terminal equipment (TE). TE type 1, i.e., TE1,
and TE type 2, i.e., TE2. The TE1s are specialized ISDN terminals to connect to
the ISDN network through four-wire, twisted-pair digital links, while the TE2s
connect to the ISDN network through a terminal adapter (TA).
● TAs: An ISDN TA can be either a standalone device or an integrated board within
TE2, which is basically used to connect the ISDN to other interfaces, e.g., serial
interface on a router.
● Network-termination (NT) devices: also include two NT types, i.e., NT1 and NT2,
corresponding to TE1 and TE2 to connect the four-wire subscriber wiring of TE
to the conventional two-wire local loop at customer sites.
● Line-termination equipment: is a device physically connected to the phone
company switch.
● Exchange-termination equipment: is a device used to connect between the
switches within the phone network.
Switched network
NT1/NT2
NT1/NT2 ISDN
ISDN switch
Packet network
switch
TE
TE
Private line network
Figure 1.33 Integrated Services Digital Network (ISDN) devices

DCE DCE
PSE PSE
DCE DCE
PSE PSE
DTE DTE
X.25 Network
Figure 1.34 X.25 devices
ISDN provides two following types of services:

● Basic Rate Interface (BRI): offers two bearer (B) channels and one delta (D)
channel (2B + D), where the B channel carries data at the rate of 64 kbps and the
D channel is for signaling at lower rate of 16 kbps.
● Primary Rate Interface (PRI): offers more B channels with higher signaling rate
of 64 kbps to support large applications. For instance, the PRI services in North
America and Japan have 23 B channels and 1 D channel, while other parts of the
world have PRI services with 30 B channels and 1 D channel.
1.3.3.5 X.25
X.25 is an International Telecommunication Union-Telecommunication Standard-
ization Sector (ITU-T) standard protocol in WAN that describes how to connect user
devices and network devices in PSNs along with their maintenance and operation pro-
cedures. The X.25 can operate effectively and independently of the type of systems in
the network. In order to ensure reliable communication among network devices, the
X.25 contains both SVCs and PVCs (cf. Section 1.3.3.3) within the physical circuit.
X.25 devices include data TE (DTE), data circuit-terminating equipment (DCE),
and packet-switching exchange (PSE) (cf. Figure 1.34).
● DTEs are simply end devices, e.g., terminals, PCs, or network hosts.
● DCEs are communication devices, e.g., modems and switches, that act as an
interface to link the DTEs at the ESs and the PSE at the carrier provider to
transfer data across the network.
1.3.3.6 Frame relay

Frame relay is a typical WAN protocol used in PSNs with packet-switching technol-
ogy. The frame relay was originally designed to work together with ISDN to enable end
devices to dynamically share network resources. The frame relay was standardized
DCE
DCE
DCE
DCE DCE
DTE
DTE
Frame relay
Figure 1.35 Frame relay devices
by the ITU-T and is an American National Standards Institute (ANSI) standard in the
United States (US).
Similar to X.25, frame relay devices basically include DTEs and DCEs (cf.
Figure 1.35). The data communication between devices in PSN is realized via either
SVCs or PVCs along with two following techniques:
● Variable-length packets: The packets of different lengths are switched across the
VCs until reaching the destination and thus enhance the network efficiency.
● Statistical multiplexing: The statistical multiplexing is used to control the network
access of devices to efficiently allocate bandwidth in frame relay PSN.
Due to its similarity with X.25, frame relay is also regarded as a streamlined
version of X.25. However, the frame relay typically operates over WAN facilities and
consequently achieves a higher performance with more reliable connection services
when compared to X.25.
1.3.3.7 Asynchronous transfer mode

ATM is an ITU-T standard for delivering not only data but also voice and video simul-
taneously over the same medium. Similar to X.25 and frame relay, the ATM is based on
packet-switching technology and the data communication inATM is realized via either
SVCs or PVCs. However, the ATM makes further use of circuit switching to develop
a cell-switching multiplexing technology. Such asynchronous combination technique
makes the ATM more efficient than the traditional synchronous technologies.
ATM is based on cell relay which is a high-speed transmission method to transfer
information in terms of cells which are fixed-size units. Each cell has a size of
53 bytes (or octets) containing cell-header information in 5 bytes and payload in
48 bytes. Given such small size, the cells can travel faster across the network without
CSU/DSU ATM ATM

switch switch
Packet network
ATM endpoint ATM endpoint
Figure 1.36 Asynchronous transfer mode (ATM) devices
suffering the latency caused by waiting the large data packet. The ATM therefore
provides high-speed WANs and LANs.
ATM devices (cf. Figure 1.36) consist of the following:
● ATM switch: has similar role as the switches used in different technology but
specifically designed for cell relay in an ATM network.
● ATM endpoint: is basically an ES which could be routers, LAN switches,
CSU/DSU, and workstations.
1.3.3.8 Switched multimegabit data service

SMDS is a high-speed, packet-switched, datagram-based WAN technology used
for sending and receiving data over public networks, namely, public data networks
(PDNs). By using fiber-optic or copper cables, the SMDS can support high-speed
data transmission.
SMDS devices consist of the following components (cf. Figure 1.37):
● Customer premises equipment (CPE): is a TE at the customer site, which could
include end devices, e.g., terminals and PCs, and intermediary devices, e.g.,
modems, multiplexers, and routers.
● Carrier equipment: generally includes high-speed WAN switches complying with
network specifications and standards for carrier networks.
● Subscriber network interface (SNI): is the interface between CPE and carrier
equipment to provide service transparently to the customer by using SMDS
interface protocol.
1.3.3.9 Digital subscriber line

DSL technology is a WAN technology that exploits existing twisted-pair telephone
lines to convey high-bandwidth multimedia data across the network to customer sites
(cf. Figure 1.38). Depending on the employed technologies, provided services, and
SNI
SNI
MUX
Carrier equipment
SMDS
CPE CPE
Figure 1.37 Switched multimegabit data service (SMDS) devices
PSTN
ADSL
modem DSL access DSL switch
multiplexer ISP
ATM network
Figure 1.38 Digital subscriber line (DSL)
supported data rate, there are various xDSL forms, of which can be typically listed
as follows:
● Asymmetric DSL (ADSL): is the most popular associated with DSL technology.
In the ADSL, more bandwidth is allocated for downlink compared to uplink, and
thus suitable for users who prefer to download data, surf internet, watch online
movie, etc., rather than uploading or sharing data online. With the asymmetry
in ADSL circuit design dividing the bandwidth of a telephone line into multiple
channels for high-speed downstream, medium-speed duplex, and basic telephone
service, the ADSL can enable data flow for both downstream and upstream,
while guaranteeing uninterrupted basic telephone service. The ADSL is therefore
playing a crucial role and will keep promoting in telephony markets.
● Symmetric DSL (SDSL): provides symmetrically high-speed data communication
of variable rate up to 1.54 Mbps. The SDSL, though supports only data yet does
not allow analog signal as in ADSL, is easy to deploy and is thus beneficial for
the services required only high-speed data transmission over a long distance.
● High-bit-rate DSL: is a symmetric version of DSL that can offer full-rate service
of 1.5 Mbps over two-wire pairs but does not support adaptive-rate services as in
SDSL.
● Very high-data-rate DSL: enables high-speed data transmissions of nearly ten
times than that of the ADSL over short distances using twisted-pair copper
telephone lines as in ADSL.
1.4 Open systems interconnection reference model

In order to convey data between devices across different network infrastructures over
a long distance, intermediary hardware is certainly vital to connect them together.
However, how can an application in one device communicate with another application
in another device?
In 1984, the International Organization for Standardization (ISO) decided to
establish a common reference model, namely, OSI, as a conceptual framework for
communication between PCs in the network. With the OSI reference model, the
vendors and developers can design digital communication products and software
programs that can interoperate and compatible with each other.
The OSI reference model divides the process of communication between two end
devices, i.e., PCs, into seven layers, each of which specifies distinct network functions
and tasks and can be implemented independently (cf. Figure 1.39). For instance, a PC
wants to send a message to another PC. The message will be sent from an application
of the source PC go down through all the layers across the network and then go up
through the same layers at the destination PC. In this section, I will first describe seven
Data Data
Application layer
Presentation layer
Session layer
Transmit data
Receive data
Transport layer
Network layer
Data-link layer
Physical layer
Transmission medium
Figure 1.39 Open systems interconnection (OSI) layers

layers of the OSI model, followed by the introduction of the interactions between these
layers in communications as well as the information format associated with each layer.
1.4.1 OSI layers

Starting from the layer nearest to the user running various applications, the seven OSI
layers in Figure 1.39 can be described in the following sections.
1.4.1.1 Layer 7—application layer

Application layer is the OSI layer 7 at which the end users interact directly with the
application. This layer is not an application but provides services for an application
program to guarantee effective communication between applications in the network.
At the application layer, the functions and services provided involve identifying
and authenticating communication partners, assessing network resources, maintain-
ing synchronization of communications, and determining protocol for the software
application. Some examples of protocols at the physical layer include File Transfer
Protocol (FTP), Telnet, Simple Mail Transfer Protocol (SMTP), and Simple Network
Management Protocol (SNMP), which will be discussed in detail in Section 1.5.5.
1.4.1.2 Layer 6—presentation layer

Presentation layer is normally a part of an operating system (OS) to provide a vari-
ety of coding and conversion functions on the data at the application layer. The
presentation layer basically turn the data received from the application layer into
a presentation format that can be converted back at the other end using common
data representation formats, common data compression schemes, and common data
encryption schemes. Some well-known standards for text include Extended Binary
Coded Decimal Interchange Code (EBCDIC) and American Standard Code for Infor-
mation Interchange (ASCII), those for video include QuickTime and Motion Picture
Experts Group (MPEG), those for graphic image include Graphics Interchange For-
mat (GIF), Joint Photographic Experts Group (JPEG), and Tagged Image File Format
(TIFF).
1.4.1.3 Layer 5—session layer

Session layer establishes, manages, coordinates, and terminates communication ses-
sions, including service requests and responses between applications at two end
devices. The services at the session layer also involve authentication and reconnection
of the communication sessions when interrupted.
1.4.1.4 Layer 4—transport layer

Transport layer segments the data received from the session layer into packets for
transporting across the network. The services at the transport layer includes error
checking of the received data making sure that the data is delivered reliably in a
proper sequence with flow control. Multiplexing and VCs are also set up at this layer
to enable data from various applications to be transmitted between the devices on a
single physical link. On the Internet, the transport protocols areTCP and user datagram
protocol (UDP), which will be described in Sections 1.5.3 and 1.5.4, respectively.
1.4.1.5 Layer 3—network layer

Network layer manages the addressing and routing of the data packets received from
the transport layer to the right destination based on the network address. A typi-
cal example of an network layer implementation is the IP on the Internet. Detailed
description of the IP will be provided in Section 1.5.2.
1.4.1.6 Layer 2—data-link layer
Data-link layer establishes reliable data transmission across the physical link in the
form of network frames. Functions at the data-link layer include physical address-
ing, network topology formation, error notification, sequencing of frames, and flow
control. The data-link layer has the following two sub-layers:
● Logical link control (LLC) sub-layer: manages communications between devices
over a shared physical data link. The LLC is defined in the IEEE 802.2
specification.
● Media access control (MAC) sub-layer: manages protocol access to the physical
network medium via a MAC address which is a unique identifier defined by the
IEEE for a NIC of a manufacturer.
1.4.1.7 Layer 1—physical layer
Physical layer determines various mechanisms and specifications for conveying bit
streams over the network. The physical layer specifications are defined depending
on the hardware facilities, which could include voltage levels, bandwidth, maximum
transmission distance, etc.
1.4.2 Interaction between OSI layers in communications

The communications between two applications in two end devices have to go through
all seven OSI layers as defined in the above subsection, but in reverse directions.
Assume that an application in PC A wants to send information to another application
in PC B. The information flow from the application program in PC A will be passed
from the application layer sequentially through all layers to the physical layer of PC
A. In the form of bit streams, the information will be sent across the transmission
medium to PC B where the information will go through all layers from the physical
layer to the application layer. Then, the received information will be processed with
an application program in PC B, which complete the communication process.
In general, a layer in the OSI model communicates with the layer directly above
it (except the application layer), the layer directly below it (except the physical layer),
and its peer layer in the remote PC making use of the services provided by adjacent
layers. For instance, the data-link layer in PC A interacts with its adjacent network
layer and physical layer in addition to the data-link layer in PC B.
Considering the characteristics and interaction of the seven layers, the OSI model,
for simplicity, can be divided into two sets (cf. Figure 1.40):
● Upper OSI layers: include the application, presentation, and session layers to
manage the information process related to the software.
● Lower OSI layers: include the transport, network, data-link, and physical layers
to control the data delivery until reaching the transmission medium.
Data
Application layer Layer 7

Upper
Presentation layer Layer 6 OSI
layers
Session layer Layer 5
Transport layer Layer 4
Network layer Layer 3 Lower

OSI
Data-link layer Layer 2 layers
Physical layer Layer 1
Figure 1.40 Upper and lower OSI layers
1.4.3 Information exchange and information formats in OSI layers

For information exchange between PCs in different systems or networks, control
information is required in the OSI layers (cf. Section 1.4.1). The control information
is defined in variety of forms to exchange between peer layers. Specifically, either
headers or trailers or both are prepended to the data information when it passes down
from upper layers. Such process is called encapsulation where the data is enclosed
with the control information to allow its readability or decodability at the other end
device. In reverse, decapsulation is performed at the destination PC to sequentially
remove the appended headers and/or trailers.
Through the encapsulation process, there are various information formats, in
which the common ones are as follows:
● Frame: is an information unit at data-link layer, consisting of the data-link layer
header and/or trailer and data encapsulated at the network layer.
● Packet: is an information unit at network layer consisting of the network layer
header and/or trailer and data encapsulated at the transport layer.
● Segment: is an information unit at transport layer employing TCP protocols with
packet segmentation. A segment consists of the transport layer header and/or
trailer and data encapsulated at the session layer.
● Datagram: has the same format as segment at transport layer but associated with
UDP protocols.
● Message: is an information unit often at application layer.
● Cell: is a fixed-size information unit at data-link layer but is specially used in
PSNs with ATM (cf. Section 1.3.3.7) and PDNs with SMDS (cf. Section 1.3.3.8).
Besides the above formats, there exist a general term, namely, data unit, to
represent information units at different OSI layers. Common data units consist of
service data unit (SDU) and protocol data unit (PDU). The SDU is a unit of data
passed down from an upper layer to a lower layer that has not been encapsulated, while
Data
PDU
Application layer
Presentation layer Data
Session layer
Segment/
Transport layer
Datagram
Network layer Packet
Data-link layer Frame
Physical layer Bit
Figure 1.41 Protocol data unit (PDU) in OSI model
the PDU refers to a unit of data after encapsulating the SDU with control information.
The PDUs are defined according to the OSI model layers (cf. Figure 1.41). In the
upper OSI layers, i.e., application, presentation, and session layers, PDU is referred
to as data. In the lower OSI layers, there are four PDUs as follows:
● Layer 4—transport layer: PDU is a segment (TCP)/datagram (UDP).
● Layer 3—network layer: PDU is a packet.
● Layer 2—data-link layer: PDU is a frame.
● Layer 1—physical layer: PDU is a bit.
1.5 Internet Protocols

Although the OSI reference model is well described through seven layers (e.g., Section
1.4), it is not a communication method, yet a conceptual framework for commu-
nication between two end devices in the network. An actual data transmission or
communication is only enabled and activated through a communication protocol
which is a set of rules that regulate the exchange information and make the communi-
cation more efficient over a network medium. There are a variety of communication
protocols operating at different layers in LANs and WANs. This section will discuss
the most widely used protocol suite used by the Internet called IPs, of which the
most communication protocols are the TCP and the IP. In order to find an appro-
priate way for the packets to travel from the source to the destination, it is crucial
to have routing protocols. Various routing protocols will be explained in detail in
Chapter 2, while the rest of this section will outline other protocols, including UDP,
Internet Control Message Protocol (ICMP), Address Resolution Protocol (ARP), and
Data OSI model TCP/IP suite
Application layer
Presentation layer Application layer
Session layer
Transport layer Transport layer
Network layer Internet layer
Data-link layer
Network access layer
Physical layer
Figure 1.42 Transmission Control Protocol (TCP)/Internet Protocol (IP)
application-layer protocols, such as FTP, Trivial FTP, HyperText Transfer Protocol

(HTTP), Telnet, SNMP, SMTP, and domain name system (DNS).
1.5.1 TCP/IP suite

TCP/IP suite2 is a dominant commercial protocol architecture for the Internet and
the WWW. Invented by Vinton G. Cerf and Robert E. Kahn in 1974, the TCP/IP
was indeed specified and extensively used 10 years before the proposal of the OSI
framework by the ISO. The TCP/IP has been developed by the US Defense Advanced
Research Projects Agency and has been regarded as the heart of the Internet providing
fundamental communication protocols until now.
Although there is no official model for the TCP/IP suite, it is really a working
one with the following layers (cf. Figure 1.42):
● Application layer: provides network services to applications with many appli-
cation protocols, e.g., HTTP, FTP, Telnet, SNMP, SMTP, DNS, NFS, XDR,
RPC.
● Host-to-host or transport layer: deals with the transmission of the data with two
main protocols including TCP and UDP.
● Internet layer: incorporates how the packet is constructed for transmission,
represented by the IP.
● Network access layer: encapsulates IP packets into frames for transmission, map
IP address to physical hardware address, i.e., MAC address, and transmit data
physically over the medium.
The above layers of the IPs spread across all seven layers of the OSI reference model
with some similarities yet simplicity with only four layers.
2
TCP/IP suite can also be referred to as the Internet protocol suite.
1.5.2 Internet Protocol

IP is a protocol at Internet layer of TCP/IP suite, which is corresponding to network
layer in the OSI model. The IP contains packets which are blocks of data to be directed
over networks using routing protocols. An IP packet basically include a source IP
address, a destination IP address and the actual data to be conveyed along with some
control information.
1.5.3 Transmission Control Protocol

TCP is a protocol at transport layer of TCP/IP suite, which has the same name as in the
OSI model. The TCP is designed to help the IP to provide reliable data transmission.
The TCP is a connection-oriented communication protocol supplying a VC between
end-user applications at source and destination devices. The TCP divides outgoing
messages into segments and then passes them to the IP for routing to the destination
where the messages will be reassembled from the incoming segments. In order to
provide reliable communications, the TCP also allows the devices to resend the lost
packets and erroneously received packets through an acknowledgment mechanism.
Several well-known application-layer protocols that use TCP include FTP, HTTP,
Telnet, and SMTP.
1.5.4 User datagram protocol

UDP is also a protocol at transport layer of TCP/IP suite as in the TCP. However, the
UDP is designed for connectionless communications. The UDP is considered as an
unreliable protocol since there is no acknowledgment mechanism for retransmission
of the lost packets. The UDP also provides no software checking for the segment
delivery and thus cannot guarantee the packet delivery. Although the UDP seems to
have no benefit over the TCP, such lack of control information for simple communica-
tion makes the UDP much faster than the TCP. The UDP is therefore more appropriate
for time-sensitive and real-time applications, e.g., multimedia, video streaming, and
broadcasting, where the latency of the packet delivery is a major concern.
Several well-known application-layer protocols that use UDP include DNS,TFTP,
and SNMP.
1.5.5 Other protocols

There are a variety of protocols corresponding to applications and services at different
layers, which can be listed as follows:
● ICMP: is a network-layer IP that is designed to report errors of the message

transmission.
● ARP: is a communication protocol that is used to map IP network addresses to
hardware addresses, MAC addresses, corresponding to data-link layer in the OSI
model. Inversely, the IP addresses can be obtained from the MAC addresses using
Reverse ARP.
● Application-layer protocols: include some widely known protocols at different

port numbers as follows (sorted in ascending order of the port number):
– FTP: is a TCP used to handle data transfer on port 21.
– Telnet: is a TCP used to manage network devices at the command prompt
level on port 23.
– SMTP: is a TCP used to provide electronic mail services on port 25.
– DNS: is a UDP used to translates domain names of the network into IP
addresses on port 53.
– Trivial FTP (TFTP): is a UDP used to convey data on port 69 in connectionless
communication with no requirement of session establishment and release.
– HTTP: is a TCP used by web browsers on port 80.
– SNMP: is a TCP/UDP (port 161/162) used by network administrators to mon-
itor and report abnormal network conditions, configure, and control network
devices.
1.6 Internetworks
As shown in previous sections, there exists a variety of network models with different
sizes and functionalities. It is vital to connect them together in a large network, namely,
internetwork, by using intermediary devices.
When considering a large group with a number of individual elements, one of the
questions that can be raised is how to share a limited resource available in the group
to every element in an efficient way. This concern indeed applies to the internetworks
which also require an efficient resource sharing and effective communication capa-
bility between all individual networks. In other words, network management plays an
important role in internetworks to cope with the incompatibility of different devices
as well as the inconsistency of different network models.
With the evolved technologies, many manufacturers, vendors, and operators are
continuing to create new machines with different configurations, OSs with differ-
ent programs, and also different transmission media with advanced communication
mechanisms. Internetworking therefore has to enable the compatibility of them allow-
ing the cooperation of PCs from different vendors with different OSs using different
transmission media on different networks in different continents.
Internetworking is not simply providing a connection between devices to link
them together. Such process sounds simple, but it in fact requires complicated tasks
with challenges to be faced. There are a number of concerns regarding the connectivity,
compatibility, flexibility, reliability, and security in the internetworks. For instance,
one of the critical issues is how to provide a reliable and secured communication
between different devices over different media in different systems given a number
of advanced technologies evolved along with cyberattacks developed in parallel, e.g.,
distributed denial-of-service attack and recently WannaCry ransomware attack have
affected computers worldwide.
In order to easily manage such a large network, internetworks require a network
hierarchy for organizing devices in systems and an addressing mechanism to identify
AS
ES IS ES
IS
Area
Area
IS IS
ES
ES IS
IS
ES ES
Area
Area
IS
Figure 1.43 Internetwork hierarchy
these devices following international standards. In the following, internetwork hierar-

chy and addressing will be described, followed by the introduction of some well-known
standards organizations and intermediary devices used in the internetworks.
1.6.1 Internetwork hierarchy

For ease of management, the ISO has defined key terminologies for different systems
as follows (cf. Figure 1.43):
● ES: consists of PCs, printers, terminals, etc. at the end users who only perform
tasks with applications.
● Intermediate system (IS): consists of repeaters/hubs, switches/bridges, and
routers, which are used to connect PCs in either the same or different networks.
● Autonomous system (AS): is a “large” network consisting of a number of networks
interconnected via routers and controlled by a common administrative entity. Due
to its large size, the AS is normally divided into subnetworks or areas depending
on their location and routing policies in the network.
1.6.2 Internetwork addressing

To identify devices in the internetworks, there are basically two internetwork address
types, including
1.6.2.1 MAC addresses

Each LAN interface has a unique MAC address assigned by the manufacturer. A PC
in the ES has only a LAN interface with only a MAC address, while internetworking
MAC address
Manufacturer’s Interface
identification number serial number
24 bits 24 bits
Figure 1.44 MAC address
devices in IS have multiple LAN interfaces for interconnection and thus should have
multiple MAC addresses. Since MAC is a sublayer of data-link layer in the OSI
framework (cf. Section 1.4.1.6), the MAC addresses are also referred to as a subset
of data-link layer addresses.
A MAC address consists of 48 bits written in hexadecimal (or base-16) numbers
(cf. Figure 1.44). For simple representation, there are two formats of MAC addresses,
i.e., MM-MM-MM-SS-SS-SS or MMMM.MMSS.SSSS. The first 24 bits are identi-
fication number of the manufacturer and the remaining 24 bits are the interface serial
number assigned by the manufacturer.
Example 1.1. A MAC address 00000ABC1982 can be formatted as
● 00-00-0A-BC-19-82
● 0000.0ABC.1982
1.6.2.2 Network layer addresses

A network layer address is used at the network layer in the OSI model (cf. Section
1.4.1.5) as virtual or logical addresses. The network layer address is required for
network layer protocols, i.e., routing protocols, to determine the location of devices
following the internetwork hierarchy model (cf. Section 1.6.1). Depending on the
number of physical network connections at the internetworking devices, they may
have one or multiple network layer addresses.
Over the Internet with IP routing protocol, each device has a logical address
called IP addresses. There are two versions of IP addresses, namely, IP version 4
(IPv4) and IP version 6 (IPv6). An IPv4 address consists of 32 bits represented in
dot-decimal format as A.B.C.D where A, B, C, and D are in decimal representing
eight bits called an octet. An IPv6 address is much longer including 128 bits, i.e., 16
octets, represented in hexadecimal numbers and separated by colons for every two
octets. The IP addressing will be discussed in detail in Chapter 3.
Example 1.2. IPv4 and IPv6 addresses:
● IPv4: 172.10.10.82
● IPv6: 1982:BA10:C12:7:10:14:12:30
1.6.3 Standards organizations

Standards are required to allow for interoperability between devices in different
networks. The standards are helpful in not only ensuring a large market for equip-
ment from different manufacturers, but also enabling the communications between
the products from different vendors. However, the standards may cause freezing
technologies, and there may have multiple standards for the same things.
For the standard development, prior to putting the standards into formal spec-
ifications, organizations provide forums and online societies for discussion, e.g.,
Internet Society, ATM Forum, Internet Activities Board with Request For Comments
documents. Some well-known standards organizations can be listed as follows:
● ISO
● International Telecommunication Union—Telecommunication Standardization
Sector (ITU-T)—formerly Consultative Committee for International Telephony
and Telegraphy (CCITT)
● European Computer Manufacturing Association (ECMA)
● Electrical Industries Association (EIA)
● Telecommunication Industries Association (TIA)
● American National Standards Institute (ANSI)
1.6.4 Internetworking devices

To enable data communications between networks or areas within an AS, internet-
working devices are required to connect:
● LAN-to-LAN
● LAN-to-WAN
● WAN-to-WAN
There are four primary types of devices, including:
● Repeaters/hubs
● Bridges/switches
● Routers
● Gateways
A quick glance on the above list is that some of them except gateways have the same
names as those discussed in Section 1.2.1 when I briefly introduced intermediary LAN
devices. Indeed, those devices, i.e., repeaters, hubs, bridges, switches, and routers,
are also internetworking devices having the same functions as in LAN. With reference
to the OSI layers (cf. Section 1.4.1), the internetworking devices can be classified as
● Repeaters/hubs: OSI layer-1 devices at the physical layer.
● Bridges/switches: OSI layer-2 devices at the data-link layer.
● Routers: OSI layer-3 devices at the network layer.
● Gateways: operate at or above OSI layer 4, i.e., from the transport to application
layer.
This subsection will discuss more details of the devices in the internetworks.
1.6.4.1 Computers
Along with the intermediary devices in the internetworks, there is a must-have device
that every computer system cannot run without it. That is computer or PC or terminal.
Every PC has a unique way to identify itself, which is based on a physical address.
A PC has a physical address regardless it is connected to internetwork or runs on its
own. The physical address can be found on the NIC which is a circuit board or a chip
allowing a PC to communicate with other PCs in the network.
The physical address is programmed into a chip on the NIC and is assigned by the
hardware manufacturer before it leaves the factory. Since the physical address is also
referred to as MAC address (cf. Section 1.6.2.1), the NIC is a layer-2 device at the
data-link layer. The address is unique on every NIC card, so if the NIC were replaced
on a PC, the physical address of the PC would change to that of the new NIC.
1.6.4.2 Repeaters/Hubs
A repeat, as the name says, is basically an intermediary device used to extend the
coverage of LANs. It has two ports and the data flow can travel in both ways, i.e.,
two-way communications. When receiving the data from one port, the repeaters may
clean up the noise, amplify, and pass the data at bit level to another port; hence, the
repeaters function at the physical layer of the OSI framework. The repeaters therefore
depend on the transmission media, but they are independent of the communication
protocols. Affected by the transmission media, the repeaters are normally confined
to a house or a single building.
The repeaters provide no traffic isolation of all data flows from the computers
attached to them in the network. The monitoring of the operation at the repeaters
is limited, and thus using the repeaters should be aware of the connection links,
especially those which are likely to fail. The number of repeaters in the network is
also restricted due to their limited transmission distance.
Following are the two common types of repeaters:
● Simple repeaters: are simply used to get around limitations in cable length caused
by signal loss or timing dispersion. They pass on individual data bits without
performing any processing at the packet level. The collisions are also passed
along with the data with no restriction. Therefore, the number of the repeaters
that can be used in the network is limited and the basic Ethernet design requires
that signals must be able to get from one end of the network to the other within a
specified amount of time and a maximum allowable length.
● Buffered repeaters: operate at the level of the whole data packets. Rather than
passing bit by bit as in the simple repeaters, the buffered repeater processes an
entire packet received from a network, stores the packet in an internal buffer,
and then retransmits it at the other end port to another network. The collisions
are not passed and repeated, and thus, there is no restriction of the number of
the buffered repeaters that can be used in the network. There is only a require-
ment that the packets received from two networks should have the same format,
while the requirement that both the networks should be of the same type can be
relaxed.
As an extension of repeaters, a hub is simply regarded as a multi-port repeater.

The hub is a very common internetworking device and is used as the center of a star
topology network.
1.6.4.3 Bridges/Switches
A bridge, like a repeater, is used to store and transmit packets and also dependent on
the transmission media. However, the bridge functions at the data-link layer of the
OSI model. The bridge cannot recognize the protocols used at the layers above the
data-link layer, but it simply passes all protocols onto the network.
One of the advantages of the bridge is that it can isolate traffic by dividing
the network to reduce traffic load on the network. Therefore, more repeaters can
be used. Operating at the data-link layer, the bridges can determine whether the
destination MAC address carried by data is not part of the same network segment
as its source; however, it cannot determine what network segment the data should
be sent to. In that case, the bridges pass randomly data across all segments of the
network. Such broadcasting may cause broadcast storms, which is a major drawback of
the bridge.
With multiple ports, a multi-port bridge is named as a switch and is popularly
used in internetworks.
1.6.4.4 Routers
Considering a large network including a number of network segments or subnetworks
using different protocols, hubs and even switches are shown to be not able to support
all of them. In such a complex network, it is required to have an intermediary device
that can find the best path for the data transmission between the segments based on
their address. A router was therefore designed.
The routers operate at the network layer of the OSI framework. Different from
the hubs and switches, the routers are independent of the transmission media. The
routers are generally much more complex and expensive than hubs and switches,
and thus they are used only when routing is required to connect networks having
dissimilar protocols. The routers make their decisions on the path routing based on
network layer address (cf. Section 1.6.2.2) rather than the destination MAC address
(cf. Section 1.6.2.1). In Chapter 2, routing protocols will be discussed in detail.
1.6.4.5 Gateways
Basically, a gateway can be viewed as a general version of a router, which also has
a basic function to connect different networks. However, the gateways operate at
the transport layer or above layers in the OSI model. The gateway can convert data
format from the application program, languages, architecture, and communication
protocols at a system to make them conform with others at the destination system.
The gateway is therefore more complex than the router in enabling the communication
and understanding between different architectures and environments.
1.7 Related works

For more information, the readers are suggested to refer to the following selected
references:
Data Networking Spohn et al. in 2002 [1], Stallings in 2006 [2], and Dye et al. in
2011 [3].
LAN Fraser in 1983 [4], Chu and Gans in 1987 [5], Pahlavan et al. in 1995 [6],
Crow et al. in 1997 [7], Woodcock in 1999 [8], Clark and Hamilton in 1999 [9],
Bray and Sturman in 2000 [10], Stallings in 2001 [11], and Roshan and Leary in
2003 [12].
WAN Decina and Scace in 1986 [13], Lechleider in 1991 [14], Thorpe and Ross in
1992 [15], McDysan and Spohn in 1994 [16], Brown and Malis in 1998 [17],
Ginsburg in 1999 [18], and Bellamy in 2000 [19].
OSI Model ISO/IEC standard 7498-1 in 1994 [20].
TCP/IP Cerf and Khan in 1974 [21] and Rodriguez et al. in 2001 [22].
Internetworks Boggs et al. in 1980 [23], Teare in 1999 [24], and Stallings in
2003 [25].
1.8 Review questions

[Q1.1] List popular computer network models and explain how to differentiate
between them.
[Q1.2] Differentiate between hubs, switches, and routers in a LAN.
[Q1.3] How are the devices physically connected in a LAN? Discuss advantages
and disadvantages of each topology.
[Q1.4] Compare different technologies used in a LAN.
[Q1.5] What are the differences between a WLAN and a LAN, between Wi-Fi and
Bluetooth, and between token ring and Ethernet.
[Q1.6] Which access methods are employed in the IEEE 802.3, IEEE 802.5, and
IEEE 802.11? Discuss and explain the reasoning for the selected methods
in each standard.
[Q1.7] Name three transmission methods in a LAN and give an example for each
of them.
[Q1.8] List some common devices used in a WAN and discuss their usage.
[Q1.9] Differentiate between full mesh and partial mesh topologies. Which one is
feasible in a large network?
[Q1.10] What are the differences between circuit switching, datagram packet
switching and VC packet switching? Discuss the advantages and disad-
vantages of them for data transmission.
[Q1.11] Differentiate between SVC and PVC in terms of resource requirement.
[Q1.12] What services are provided by ISDN and what data rates are offered by
each service?
[Q1.13] What are the similarities and differences between X.25, Frame relay, ATM,
and SMDS?
[Q1.14] Describe some typical forms of DSL and their applications.
[Q1.15] List seven layers of the OSI reference model in order and the corresponding
PDU at each layer.
[Q1.16] Describe TCP/IP suite with its layers and discuss the necessity of layering
with the TCP/IP.
[Q1.17] Differentiate between TCP and UDP.
[Q1.18] List and briefly describe some application-layer protocols usingTCP and/or
UDP.
[Q1.19] What are internetworks?
[Q1.20] How to identify devices in the internetworks?
[Q1.21] Describe briefly intermediary devices used in internetworks and discuss
how to differentiate between them.
1.9 Problems
[P1.1] Consider a network consisting of N devices. How many cable links are
required in total and how many ports are required in each device if they are
connected in a ring, bus, star, and full mesh topology? Illustrate the answer
with five devices.
[P1.2] Sketch a hybrid topology to connect 40 devices with a star backbone and
four ring networks. Modify the topology using a ring backbone and two bus
networks.
[P1.3] Given six devices with ten cables, design a partial mesh network using all
cables.
[P1.4] Consider a PCN employing VC to connect node A and node B via three
nodes 1, 2, and 3. The length of the message to be sent is 40 octets and
every packet contains a header of 3 octets for control information. Assume
that it takes 1 ms to send an octet, and there is no delay caused by the
processing at the nodes.
● Determine the transmission time to send a message from A to B.
● Suppose that the message is split into two packets. Find the new trans-
mission time from A to B. Discuss the findings when the message is
split further into four or five or ten packets.
● Let N , K, L, and H denote the message length in octets, header length in
octets per packet, the number of packets in a message, and the number of
hops between two end devices, respectively. Assume that N is divisible
by L. Derive the expression of the transmission time between these two
end nodes, which is also known as end-to-end delay.
● Find the optimal L, as a function of K, L, and H , so as to minimize the
end-to-end delay in VC-based PCN.
[P1.5] Derive the general expression of end-to-end delay in a network employing

either circuit switching or datagram packet switching orVC packet switching
as a function of the following parameters:
● N : message length in octets
● K: header length in octets per packet
● H : number of hops between two end devices
● L: number of packets per message
● R: link data rate in bps
● Ts : call setup time, in seconds, for circuit switching and VC
● Tp : propagation delay over a hop in seconds per hop
Find the optimal L, as a function of the other parameters, to minimize the
end-to-end delay in datagram packet switching.
Chapter 2
Routing in computer networks
In an internetwork with a number of network segments using a variety of protocols

and network topologies, the communications between devices in these subnetworks
require a powerful device to connect them in an efficient way. It is therefore vital
to have a router as an intermediary device that not only plays the role of con-
necting the subnetworks but also routes the data via an optimal path toward the
destination node.
Routing is basically finding the path between the source and destination nodes
based on their addresses. The work of the router sounds like that of switches; how-
ever, the routers operate at the network layer of the OSI framework, while the switches
function at the data-link layer using MAC addresses. The routers are independent of
transmission media and are generally much more complex requiring high invest-
ment expenditure compared to other intermediary devices at lower layers. In order
to route the messages between networks, a routing protocol is crucial, of which path
determination is the most complicated process.
In this chapter, for a first view of routing in computer networks, I will first dis-
cuss in Section 2.1 the fundamentals of routing along with an overview of algorithms
for path determination. Static versus dynamic routing protocols will be described
in Sections 2.2 and 2.3, respectively. Considering the routing protocols in different
network infrastructures, interior gateway protocols (IGPs) and exterior gateway pro-
tocols (EGPs) will be sequentially presented in Sections 2.4 and 2.5 with relevant
protocols, such as Routing Information Protocol (RIP), Interior Gateway Routing
Protocol (IGRP), Open Shortest Path First (OSPF), Enhanced IGRP (EIGRP), and
Border Gateway Protocol (BGP).
2.1 Fundamentals of routing

As briefly outlined in Section 1.6.4.4 of Chapter 1, routers can make their decisions
on the path from a sender to a receiver based on the network addresses, but how do
they work?
The routers need a routing protocol which has enough network layer addressing
information for user traffic to be directed from one network to another one. There
are basically two activities involved in the routing process, including path determina-
tion and packet switching. While the packet switching requires a simple algorithm to
examine the address of the next hop for the packets to be forwarded, the path deter-
mination, which will be explained in the following subsection, is rather complicated
with different routing protocols.
2.1.1 Path determination

In fact, a routing protocol is not simply finding a path, but it should be in an efficient
way to carry messages in terms of packets between networks given a number of
available paths. The routing protocols define the format of some fields within a packet
as a preliminary information for the path determination. With the routing protocols,
it should be noticed that they are used to exchange information between routers and
do not take into account any user data and traffic.
In order to determine which path for the packets to travel, routing protocols
define route metrics which are the costs of reaching a destination. The costs here
could be the number of hops or other measurement of communications, such as delay
and bandwidth. For path determination, there are routing tables containing route
information in the network which can be regarded as an abstract of the real network
topology. The routing tables are managed by routers and are used in the routing
algorithms to calculate the optimal path to a destination. Depending on the routing
algorithms, there may have different route information stored in these tables. When
receiving a packet, the routers check its destination address and route it to the next
hop following the optimal path. The routers are therefore required to communicate
between them to keep updating their routing tables.
2.1.2 Routing algorithms

There exist various routing algorithms depending on network resources, router capac-
ity, and route metrics as well as practical requirements. The design of all these routing
algorithms however has the same desired outcome which is the best or optimal route(s)
from one point to another point. These are required to be stable but robust to be able
to confront the hardware failures in unpredictable circumstance that may cause load
unbalancing, overburden, and even interruption. Moreover, the routing algorithms
should be designed in an efficient way with low complexity and low overhead to
allow them function in the condition of limited physical resources. The convergence
of the algorithms implemented at the routers is also a critical design goal to guar-
antee that the best route can be determined after a limited number of routing update
messages, and thus the routing loops or network outages can be refrained from the
slow network convergence. With a number of existing network infrastructures with
evolved technologies, the routing algorithms should also be adaptable to different
network models and settings along with internetworking devices and terminals with
different configurations.
Routing in computer networks 47
With different design goals, the routing algorithms are classified based on dif-
ferent criteria, such as network infrastructures, hierarchies, complexity, convergence,
adaptability, and flexibility. They may be contrary in meaning and functionality but
can supplement each other if appropriately used. For instance, some can be listed as
follows:
● Flat routing algorithms where all nodes are regarded as peers of the same level
versus hierarchical routing algorithms which are based on the network hierarchies
as modeled in an organization (cf. Section 1.6.1).
● Single-path routing algorithms for finding only one best path from source to
destination versus multipath routing algorithms which support multiple paths to
the same destination for load sharing.
● Distance-vector routing algorithms which simply use the distances between
neighboring nodes with low complexity versus link-state routing algorithms
which rely on the routing updates of the entire network and thus require high-
complexity processing at the routers but allow fast convergence.
● Static routing algorithms of which the routes are manually preset by a net-
work administrator irrespective of possible succeeding network changes versus
dynamic routing algorithms which automatically find alternative routes as the
network topology or traffic changes.
Basically, with respect to a specific router in a network, there are two kinds of
network connection which are either directly or remotely connected. The associated
networks of the router can be therefore categorized as follows:
● Directly connected networks: are networks connected directly with the router
interfaces via directly connected routes.
● Remote networks: are networks connected through other routers by using routing
protocols with either static or dynamic routes. Details of the static and dynamic
routing protocols will be presented in Sections 2.2 and 2.3, respectively.
The routing algorithms can be either implemented within an enterprise or across

the Internet. The routing protocols can be accordingly grouped into following two
groups (cf. Figure 2.1):
● IGP: is used to describe a routing protocol employed within an autonomous

system (AS) in the internetwork hierarchy. Depending on the employed rout-
ing algorithms, there are different forms of the IGP, which will be discussed in
Section 2.4.
● EGP: is a routing protocol used for exchanging routing information among ASs
on the Internet. The EGP is however obsolete after its third version and has been
replaced by a new standardized protocol called BGP. Two forms of the BGP,
including Internal BGP (IBGP) and External BGP (EBGP), will be described in
Section 2.5.
AS 1
IGP EGP AS 2
IGP
EGP
AS 3
EGP
IGP
IGP: RIP, IGRP, EIGRP, OSPF
EGP: BGP
Figure 2.1 Interior gateway protocol (IGP) versus exterior gateway protocol
(EGP)
2.2 Static routing protocols
A static routing protocol can be identified as bus routes based on which the bus uses
the same route everyday to travel between two end stations through the same number
of stops. In the context of networking, the bus can be regarded as a packet, whereas
the intermediary stops are routers in the network.
The static routes are manually defined by system administrators and are fixed
regardless of network changes or reconfiguration. These routes are easy to configure
and maintain in a small network, but it will be time-consuming with a considerably
increased complexity as the network size increases. The static routing is therefore
suitable for only a network with simple topologies, a small number of nodes, and
when the links are unlikely to change. Since the administrators are required to con-
figure the routes, their intervention is indispensable whenever the network topology
changes. The static routing protocols are however shown to provide more secure
communications with no extra resources for traffic reduction and also with minimal
processing.
2.3 Dynamic routing protocols

Unlike static routing, dynamic routing protocols allow the packets to find a new path
to the destination in an adaptive manner, and thus the dynamic routing is also known
as adaptive routing. This fact is related to the circumstance when driving a car with a
number of alternative routes depending on different factors, such as traffic conditions,
weather, etc.
In computer networks, the adaptation property of dynamic routing protocols helps

find substitute routes as the network topology or traffic changes, such as the loss of
nodes and the interruption of connection links. The packets can be automatically
rerouted around the broken links and the failure nodes via the next optimal path when
the previous one is not available. The dynamic routing is independent of the network
size, and thus more scalable and suitable for complex network topologies. However,
there are some disadvantages of such dynamics when the network is less secure and
substantial resources are required to manage the routing process.
2.4 Interior gateway protocols

An IGP is a general term to describe a routing protocol employed within an AS
for exchanging routing information between gateways. Depending on the routing
mechanisms and metrics, the IGPs can be divided into two groups, including distance-
vector routing protocols (DVRPs) and link-state routing protocols (LSRPs), which
will be described in the following subsections. In addition, there exist hybrid routing
protocols having the features of both the DVRPs and LSRPs, of which a typical
example is EIGRP designed by Cisco.
2.4.1 Distance-vector routing protocols

DVRPs are protocols based on the distance and direction between nodes in the net-
work. The DVRPs use either the Bellman–Ford algorithm to compute the shortest
paths from a source vertex to all other vertexes in a weighted digraph or Ford–
Fulkerson algorithm to compute the maximum flow or diffusing update algorithm
full-state machine (DUAL FSM) specially developed for EIGRP.
The DVRP requires that a router informs its neighbors of topology changes peri-
odically by passing the routing table to neighbor nodes in the network. The cost of
reaching a destination is quantified using route metrics which are stored in a distance
vector in the routing table. Employing various algorithms, a router can calculate the
direction and distance to other routers.
Each router does not have any knowledge about the full network topology. A router
advertises its distance vector to other routers and also receives the advertisement of the
distance vectors from the other routers. These advertisements are then accumulated
in the routing table at each router.
Without the requirement of an actual map of the network topology with hierar-
chical design, the routing update process in the DVRP is simple and does not require
much knowledge of the administrators, but its main disadvantage is slow convergence
and may cause routing loops in some cases.
Two typical examples of DVRPs are as follows:
● RIP, RIP Version 2 (RIPv2), and RIP next generation (RIPng): where the route
metric is the hop count.
● IGRP: which is based on a composite metric including bandwidth, delay, load,
reliability, and maximum transmission unit (MTU).
2.4.1.1 Routing Information Protocol

As one of the oldest and simplest DVRP, RIP was originally specified in the RFC
1058 published in 1988 and has been most widely used on the Internet for exchanging
routing information among gateways.
In the RIP, the hop count is used as the metric for path selection. The maximum
allowable hop count is 15 hops and the routing updates are broadcast every 30 s by
default setting in the update timer of the RIP. Apart from the update timer, the RIP
has other specific timers including invalid timer, flush timer, and hold-down timer.
Although the RIP is extensively used in different networks due to its low complex-
ity, it has some limitations as the number of hops is restricted and slow convergence.
The routes in the RIP will be dropped if the number of hops exceeds 15 hops. In
other words, the destination is considered as unreachable when the metric value in the
routing table is 16. This limitation however helps prevent routing loops from counting
to infinity, which is also known as count-to-infinity problem. The routing loops will
be explained in detail in Section 2.4.1.3.
The RIP has three versions as follows:
● RIP version 1 (RIPv1): is the original version of the RIP which was defined
in RFC 1058. The RIPv1 supports only subnets of the same class having the
same length or fixed length, while it does not support different-sized subnets
with variable length subnetting since there is no subnet information carried in the
routing updates. The RIPv1 is thus also known as a classful routing protocol. In
particular, there is no router authentication in the RIPv1, which makes it exposed
to the possibility of being attacked.
● RIP version 2 (RIPv2): was originally described in RFC 1058 in 1993 and was
later standardized in RFC 2453 in 1998. The RIPv2 allows more information to
be included in the RIP packets. Specifically, an RIPv2 packet contains additional
subnet and authentication information, and thus the RIPv2, as a classless rout-
ing protocol, can support variable-length subnets and authentication mechanism.
The hop count is also limited by 15 hops and the RIPv2 maintains backward
compatibility with RIPv1. In order to reduce unnecessary load on those hosts
which are not involved in routing, instead of broadcasting the routing table as in
the RIPv1, the RIPv2 multicasts it to all adjacent routers using an IP multicast
address 224.0.0.9. Additionally, the RIPv2 packet includes route tags to distin-
guish between internal routes learned with the RIP and those learned with other
protocols.
● RIP next generation (RIPng): was defined in RFC 2080 in 1997 as an extension
of the RIPv2 to support the next generation of IP, i.e., IPv6 (cf. Section 1.6.2.2).
2.4.1.2 Interior Gateway Routing Protocol

IGRP is a DVRP developed in the mid-1980s by Cisco to exchange the routing infor-
mation between Cisco gateways or routers within an AS. The IGRP overcomes the
limitations of the RIP which has only a single routing metric, i.e., hop count, limited
by 15 hops.
Similar to the RIP, the IGRP employs distance-vector routing algorithm to deter-
mine the direction and distance to any link in the internetwork and also has four
timers consisting of update timer, an invalid timer, a hold-down timer, and a flush
timer. However, the default configurable hop count of the IGRP-routed packets is 100,
which can be configured up to 255, and the routing updates are advertised over longer
intervals of 90 s. Therefore, the IGRP can handle complex topologies in large net-
works, providing scalability and flexibility for segments having different bandwidth
and delay characteristics.
In particular, the IGRP uses a combination of variables consisting of bandwidth,
delay, load, reliability, and MTU to determine a composite metric which can be
adjusted through the use of preset constants. Here, the MTU is the maximum packet
size in bytes that a particular interface can handle and the composite metric is a
function of the bandwidth, delay, load, and reliability.
Let B, D, L, and R denote the bandwidth, delay, load, reliability, respectively.
The composite metric for the IGRP, i.e., MIGRP , is computed by

K2 B K5
MIGRP = K1 B + + K3 D , (2.1)
256 − L R + K4
where K1 , K2 , K3 , and K4 are constant metric weights used to define with the router
configuration command.
By default, K1 = K3 = 1 and K2 = K4 = K5 = 0, and thus from (2.1), MIGRP =
B + D. This means that the IGRP chooses a route based on the bandwidth and the
delay by default.
Although the IGRP has been shown to be one of the most successful routing
protocols of Cisco, it lacks of support for variable-length subnetting as in the RIPv1.
Instead of designing an IGRP version 2 as the RIPv2 with that additional capability,
Cisco decided to develop an EIGRP as an hybrid approach of both the DVRP and
LSRP, which will be discussed in detail in Section 2.4.3.
2.4.1.3 Routing loops

Routing loops are common problems in the routing update process in DVRP caused
by slow convergence. The routing loops are basically formed when there is a misun-
derstanding between the routers during the routing update and thus cause an endless
loop between the routers. This means that a packet is continuously transmitted through
the routers but could not reach its intended destination.
Example 2.1. Consider a simple network with only three routers R1, R2, and R3
(cf. Figure 2.2). A packet at R1 network is supposed be sent to R3 network through
R2. The link between R2 and R3 is down but R2 has not yet informed R1 about this
breakage. A routing loop can be formed as follows:
● R1 assumes that there is a path to R3 through R2 and transmits the data to R2.
● Since R2 knows of the broken link between R2 and R3, it tries to reach R3 via R1
and thus sends the original data back to R1.
● R1 receives the same original data from R2 and then consults its routing table.
However, since R1 has not been informed about the broken link between R2 and
Loop
R1 R2 R3
Figure 2.2 Routing loop in Example 2.1
R3, its routing table also shows that the packet can reach R3 via R2, and hence
R1 resends its data to R2; and so on.
The packet will encircle endlessly between routers R1 and R2.
With an infinite number of traffics looping to and fro between routers, the routing
loop causes a significant waste of spectrum and burdened load for the routers’ CPUs.
The routing update process will also be affected in future with more routing loops,
especially in a large network. Therefore, it is critically important to have a mechanism
to get rid of such routing loops.
There are a variety of approaches to prevent routing loops from happening, which
can be listed as follows:
● Count-to-infinity condition: is the condition when the routing updates increase
the metric value to “infinity.” Different routing protocols have different “infinity”
values. For example, RIP defines that the maximum allowable hop count is 15
hops, which means “infinity” is 16 hops.
● Hold-down timers: are used to eliminate regular update messages from improperly
restoring a bad route. The hold-down timers only allow the routers to hold any
network changes for a specified period of time. If a route is identified as likely to
be down, then any other information for that route will be ignored for a hold-down
period depending on the routing protocol. This hold-down period helps all routers
have enough time to acquire the status of the unreachable network. For example,
the hold-down timer in RIP is set for 180 s by default.
● Split horizon rule: states that a router should not advertise its routing table back
to the interface where it receives the update. The split horizon rule can be applied
together with either route poisoning to prevent a router from sending packets
through an invalid route or poison reverse to explicitly mark the route as unreach-
able in the routing update. For example, the RIP has a poisoned route with a
metric of 16.
2.4.2 Link-state routing protocols

LSRP, in contrast to DVRP (cf. Section 2.4.1), is performed by every router in the
network given the fact that the full picture or complete network topology is available
at all routers.
Specifically, every router constructs a map of the connectivity or topology of the
network in the form of a graph showing which routers are connected to which other
routers. In order to form the routing table, every router calculates the best next hop
to reach all other routers in the network rather than sharing the routing table with
only its neighbors as in the DVRP. Here, the cost of reaching a destination is also
calculated based on route metrics and only connectivity-related information, not the
whole routing table, is passed between the routers.
Given the full knowledge of the network topology, the LSRP only requires update
when there is a network change instead of performing periodic updates. The LSRP
is particularly helpful for large networks which require fast convergence; however,
it requires more CPU power and memory for processing at the routers compared to
the DVRP, which make the LSRP more expensive to implement. Also, details of the
network hierarchy are required and the administrators should have a good knowledge
of how to design a network with the LSRP.
Two typical examples of LSRPs are as follows:
● OSPF: which was designed by the Internet Engineering Task Force (IETF)
working group.
● Intermediate System to Intermediate System (IS–IS): which was designed by the
ISO for the OSI framework, not the TCP/IP protocol suite.
Although the OSPF and IS–IS were designed by different groups with different
standards, they both have many similarities in providing the necessary routing func-
tionality with link-state routing algorithm. So, only the OSPF is selected to discuss
in the following subsection.
2.4.2.1 Open Shortest Path First

OSPF is a typical LSRP in the group of IGPs operating within an AS. Designed by the
IETF working group since its establishment in 1988, the OSPF was created to over-
come the shortcoming of the RIP which was shown to be incapable of handling large
internetworks. The OSPF has been then the most widely used IGP in heterogeneous
networks.
The OSPF has the following two versions:
● OSPF version 2 (OSPFv2): was defined in RFC 2328 in 1998 for IPv4 networks.
● OSPF version 3 (OSPFv3): was first defined in RFC 2740 for IPv6 networks and
then updated in RFC 5340 in 2008.
As the name stated, the OSPF is an open protocol having specification available
in the public and it is based on Shortest Path First algorithm. Specifically, Dijkstra’s
algorithm, which will be discussed in detail in Section 2.4.2.2, is employed to find
the shortest paths between the routers in the network topology.
The OSPF recreates the exact topology of the entire internetwork as in the LSRP
through sharing the connectivity-related information, namely, link-state advertise-
ments (LSAs), among routers. The LSAs consist of information about the interfaces
at the routers, metrics used, and other variables. Then, based on the LSAs, the OSPF
can determine the shortest paths between the routers. Here, the routing metric in the
OSPF is the path cost which can be simply computed as inversely proportional to
the bandwidth or link capacity. For example, a link with a capacity of 100 Mbps is
assumed to have a cost of one, then the 10 and 20 Mbps links will have costs of ten
and five, respectively. In order to balance the traffic load dynamically, the link metric
can be more complicated with the introduction of cost factors depending on the link
availability, reliability, and latency in addition to data throughput.
With the same concept of the LSRP, the OSPF operates following network hier-
archy and can quickly detect the network changes allowing fast convergence to the
new topology. As shown in Section 1.6.1, the AS is the largest entity consisting of
a number of networks which can be grouped into different areas. These areas are
connected via area border routers (ABRs) (cf. Figure 2.3). The LSAs from all the
routers in the same area, namely, internal routers, are collected to build a topological
database for that area, which is then managed and shared between the ABRs to have
an overall picture of the internetwork. Hence, the topological databases should be
identical at all the internal routers within the same area.
Depending on the location of the source and destination within the same area or
in different areas, there are two corresponding kinds of OSPF routing, namely, intra-
area and inter-area routing. In order to distribute or share routing information between
areas, the OSPF requires a backbone area which consists of all ABRs, other attached
routers, and also other non-backbone areas (cf. Figure 2.4). All the routers having an
interface connected to the backbone area are thus named as backbone routers. These
backbone routers employ inter-area routing over the backbone topology, while the
intra-area routing is performed at internal routers given only the knowledge of their
individual area topology.
Area 0
ABR
ABR
OSPF
Area 2
Area 1
Figure 2.3 Open Shortest Path First (OSPF) routing protocol
OSPF Backbone
Area 0 Area 3
ABR
ABR
ABR
Area 1
Area 2
Figure 2.4 OSPF over backbone area

2.4.2.2 Dijkstra’s algorithm

One of the fundamental problems in connection with graphs is to find the path having
the shortest length between nodes in a graph. In 1956, Edsger W. Dijkstra, a computer
scientist, came up with an idea on how to formulate and solve this routing problem,
which was then published 3 years later and well known as Dijkstra’s algorithm.
The Dijkstra’s algorithm has a number of variants applied in different areas. In
particular, it has been widely used in the OSPF routing protocol to find the shortest
routes between a node and all other nodes to build a shortest path tree.
Given a set of nodes connected in a graph with a nonnegative weight value, i.e.,
link metric or path cost, on every edge, the Dijkstra’s algorithm assigns initial distance
values which will be then updated as in the following steps:
● Step 1: Assign a provisional distance value to every node: zero for the initial node
and infinity (∞) for all other nodes.
● Step 2: Mark all nodes as unvisited and set the initial node as the current node.
The unvisited set consists of all the unvisited nodes.
● Step 3: For the current node, consider all of its unvisited neighbors and calculate
the tentative distances from this current node to the neighboring nodes. If this
new distance is less than the previously assigned distance, then overwrite with
the new one; otherwise, keep the old one.
● Step 4: After considering all the neighboring nodes of the current node, mark the
current node as visited and remove it from the unvisited set. This visited node
will not be checked ever again and its recorded distance is now final and should
be minimal.
● Step 5: Set the unvisited node with the smallest distance (from the initial node)
as the next “current node” and continue from Step 3 until the destination node is
marked as visited.
Example 2.2. Find the shortest path between two nodes A and J in Figure 2.5.
The shortest path from A to J can be determined following Dijkstra’s algorithm as
in Table 2.1 where the bold texts in each column for each node are used to trace
back the shortest route from the initial node to that node and the bold values are the
corresponding distance value. In this example, the shortest route from A to J can be
traced back as J <- I <- G <- F <- D <- C <- A having the shortest distance of 28.
B 10 E 10 H
4 3 9 6 3 14
A 9 D 10 G 15 J
5 1 4 5 5 8
C 11 F 11 I
Figure 2.5 A graph in Example 2.2 where the value on the edge represents weight
or path cost
Table 2.1 Dijkstra’s algorithm for finding the shortest distance in Example 2.2
A B C D E F G H I J
(0) (∞) (∞) (∞) (∞) (∞) (∞) (∞) (∞) (∞)
(4, A) (5, A) (9, A) (∞) (∞) (∞) (∞) (∞) (∞)
(4,A) (5,A) (6, C) (14, B) (∞) (∞) (∞) (∞) (∞)
(6,C) (14, B) (10, D) (16, D) (∞) (∞) (∞)
(14,B) (10,D) (15, F) (∞) (21, F) (∞)
(15,F) (18, G) (20, G) (30, G)
(18,G) (20,G) (28,I)
2.4.3 Enhanced Interior Gateway Routing Protocol

EIGRP is an advanced version of IGRP (cf. Section 2.4.1.2). The EIGRP was also
designed by Cisco and was officially described in RFC 7868 in May 2016 after nearly
three decades from the development of the predecessor IGRP.
The EIGRP was evolved from the demands of the large-scale internetworks to
support classless IP addresses with variable-length subnetting. Although Cisco classi-
fies the EIGRP as a DVRP (cf. Section 2.4.1), it functions as a balanced hybrid routing
protocol integrating the capabilities of both the DVRP and LSRP (cf. Section 2.4.2).
Furthermore, the EIGRP employs DUAL to improve its operational efficiency
in finding the best path as well as alternative paths with only loop-free routes. The
DUAL was first developed in 1998 by J.J. Garcia-Luna-Aceves at SRI International to
improve the Dijkstra algorithm (cf. Section 2.4.2.2). Similar to the IGRP, the EIGRP
determines the path using a composite metric depending on bandwidth, load, delay,
reliability, and MTU (cf. (2.1)). Due to this fact, the EIGRP was considered as a
DVRP. However, in addition to the routing table, the EIGRP uses neighbor table and
topology table via DUAL, which allow the routers only send routing table data when
there is an actual change in the internetwork rather than sending the routing table
periodically. With this functionality, the EIGRP also operates as an LSRP, and thus it
is generally viewed as a hybrid routing protocol.
2.5 Border Gateway Protocols

BGP is a routing protocol employed as an EGP to exchange routing information
between ASs, for instance over the Internet with different Internet service providers
(ISPs). After a number of drafts, the current version of the BGP is version 4 (BGP-4)
which was described in RFC 4271 published in 2006.
The BGP is used to advertise network addresses between ISPs and ASs. Gener-
ally, a customer network employ an IGP, e.g., RIP, OSPF, and IGRP or EIGRP, for
exchanging routing information between routers within the network. These customer
networks are connected to the ISPs which then use the BGP to exchange routing
information between them.
AS 1
IBGP
EBGP AS 2
EIGRP
IBGP
EBGP
OSPF
AS 3
EBGP
IBGP
RIP
Figure 2.6 Border Gateway Protocols (BGPs) with Internal BGP (IBGP) and
External BGP (EBGP)
Although BGP is normally applied between differentASs and ISPs on the Internet,
it can also be used for routing within an AS to coordinate and synchronize routing
policy of routers in the AS. The BGP therefore supports both session types which are
referred to as EBGP and IBGP (cf. Figure 2.6).
The BGP is based on paths and network policies configured by network adminis-
trator to make core routing decision and thus functions as a DVRP. In order to propagate
and manage the routes, there is a mechanism for route mapping which includes a set
of rules describing different actions. However, considering a full mesh configuration
of all routers within an AS employing the IBGP, each router is required to connect to
all other routers, which causes a quadratic grow of network connections between the
routers. Therefore, as a means to alleviate this scaling problem, the BGP makes use
of a variety of route parameters, namely, attributes, to define different route policies
for route selection process, as well as implementing route reflection (cf. RFC 4456)
and AS confederation (cf. RFC 5065). These additional features enable the BGP to
meet the scalability as well as the continuing changes of the Internet infrastructure.
2.6 Related works

references:
Routing protocols Sportack and Fairweather in 1999 [26], Doyle and Carroll in
2005 [27], and Graziani and Allan in 2007 [28].
DVRP Ford in 1956 [29], Ford and Fulkerson in 1956 [30], and Bellman in 1958 [31].
RIP Hedrick in 1988 [32], Malkin and Minnear in 1997 [33], and Malkin in 1998
[34].
IGRP Rutgers in 1991 [35] and technotes of Cisco in 2005 [36].
LSRP Clausen and Jacquet in 2003 [37].
OSPF Moy in 1998 [38,39] and Coltun et al. in 2008 [40].
Dijkstra’s algorithm Dijkstra in 1959 [41].
EIGRP Garcia-Lunes-Aceves in 1993 [42], Pepelnjak in 1999 [43], and Savage et
al. in 2016 [44].
BGP Rekhter et al. in 1995 and 2006 [45,46], Halab and McPhersoni in 2000 [47],
White et al. in 2004 [48], Bates et al. in 2006 [49], and Traina et al. in 2007 [50].

[Q2.1] Describe the routing process of packets in the internetwork.
[Q2.2] Differentiate between routing algorithms.
[Q2.3] Explain the metrics used in routing protocols.
[Q2.4] What are two basic groups of IGPs? Describe the fundamental differences
between them.
[Q2.5] In RIP, which metric is used and how to determine if the destination is
unreachable?
[Q2.6] Explain the purpose of update timer in RIP.
[Q2.7] What is the limitation of the RIP?
[Q2.8] Describe three RIP versions and their applications.
[Q2.9] What is composite metric in IGRP? Explain how to determine and choose
appropriate parameters for this metric subject to specific routing demands.
[Q2.10] Explain when the routing loops happen and approaches to prevent them.
[Q2.11] Compare LSRP with DVRP.
[Q2.12] Which algorithm is employed in OSPF to find the shortest path between
nodes? Describe in detail its steps.
[Q2.13] Give a brief account of the reason why the EIGRP is referred to as a hybrid
routing protocol.
[Q2.14] Differentiate between IBGP and EBGP.
[Q2.15] How to cope with the scaling problem in the BGP?
2.8 Problems
[P2.1] What is the maximum number of nodes in a RIP network?
[P2.2] Consider the bandwidth consumption of the RIP protocol. Assume a network
consists of eight nodes and a node has three neighboring nodes.
● Estimate the number of messages exchanged per unit time.
● Estimate the size of the messages exchanged as a function of the size of
the RIP network given a RIP message consists of a 4-byte header plus
20-byte per entry and up to 25 entries per message.
● Estimate the bandwidth consumption of a RIP network.
[P2.3] Assume that there are ten routers in the network and every router has four
neighbors.
● Estimate the amount of memory required to store the information used by
the distance-vector routing given the size of each entry that RIP requires
is 10 bytes.
● Estimate the amount of memory required to store the information used
by the link-state algorithm given each entry size is 10 bytes.
[P2.4] Find the shortest distance from node A to node J in the network in Figure 2.7
using Dijkstra’s algorithm.
B 6 E
4 5
9
12
A 5 C 10 F 7 H
7
8 10
8
D 5 G
Figure 2.7 Figure for Problems 4 and 5
[P2.5] Find the shortest route from node A to other nodes in the network in
Figure 2.7.
[P2.6] Using Dijkstra’s algorithm, find which of the vertices A or B or C is the
nearest to J in Figure 2.8.
A 19 E 9 H
9 15 10 30
B 11 D 25 G 25 J
13 12 11 28
C 24 F 8 I
Figure 2.8 Figure for Problem 6

Chapter 3
Internet Protocol addressing
As discussed in Sections 1.5.1 and 1.5.2, Internet Protocol (IP) is a protocol at Internet
layer of TCP/IP suite. In order to employ a variety of routing protocols to direct
data packets over the internetwork, an address of every node in the network is of
paramount importance. The IP addressing and routing are therefore the cornerstones
of a computer network. Indeed, a packet includes not only the actual data to be
conveyed but also a source IP address and a destination IP address along with some
additional control information.
In this chapter, I will first present in Section 3.1 an overview of IP addresses with
various types and addressing mechanisms. The first version of IP addresses, namely,
IP version 4 (IPv4) will be discussed in Section 3.2 followed by various mechanisms
for dividing the network with subnetting in Sections 3.3 and 3.4 as well as how to
group the small networks via supernetting in Section 3.5. A brief introduction of the
extended version of IP addressing called IP version 6 (IPv6) will be described in
Section 3.6 along with discussions for further readings.
3.1 An overview of IP address
An IP address is a sequence of number which is assigned to each device in the inter-

network. Basically, an IP address is used to identify the host, i.e., network interface
and location, in the network. Globally managed by the Internet Assigned Numbers
Authority, the IP addresses are distributed to regional Internet registries in different
areas. These IP addresses are then assigned to every device connecting to the net-
work by a network administrator. Depending on their applications in practice, such
assignment of IP addresses can be either local or global, private or public, temporary
or persistent addressing. Given the IP addresses of all devices in the network, rout-
ing protocols (cf. Chapter 2) can be implemented allowing intermediary devices to
determine the best routes for the communications between hosts.
There are two versions of IP addresses including IPv4 and IPv6. The IPv4 was
well known as the first version of the IP developed by the DARPA, while the latter is
the latest version of the IP with a larger address space developed by the IETF.
An IP address basically consists of two main parts including network prefix and
host number. In order to identify a device in a TCP/IP network, we need to know not
only its IP address or host address but also the network address which is the prefix
of an IP address determined through a mask with bitwise AND operation. The mask
in computer networking is used as a bitmask to characterize either a network or a

subnetwork (or subnet). The network/subnet address of a host is determined by
Network/Subnet address = Host address AND (bitwise) mask (3.1)
For simplicity in representation of the network/subnet address, a slash character, i.e.,
“/”, followed by the number of bits for the network/subnet address can be used instead
of writing the mask in full. This compact representation is also known as classless
inter-domain routing (CIDR) notation.
Example 3.1. A host has an IP address 158.92.1.1 with different masks as follows:
● Network mask 255.255.0.0: The network address is 158.92.0.0 with 16 first bits
for network prefix and 16 remaining bits for host number. The IP address can be
either written as 158.92.1.1 255.255.0.0 or simply 158.92.1.1/16 (cf. Figure 3.1).
32 bits
8 bits
Host address 158 92 1 1
Network mask 255 255 0 0
Network address 158 92 0 0
Network prefix Host number
Figure 3.1 IP address with network mask in Example 3.1

● Subnet mask 255.255.255.0: The subnet address is 158.92.1.0 with 16 first bits
for network prefix, 8 bits for subnet number, and 8 last bits for host number. The
IP address can be written as 158.92.1.1/24 (cf. Figure 3.2).
32 bits
8 bits
Host address 158 92 1 1
Subnet mask 255 255 255 0
Network address 158 92 1 0
Network prefix Subnet Host number

number
Figure 3.2 IP address with subnet mask in Example 3.1

Internet Protocol addressing 63
3.1.1 Addressing types

Depending on their usage, there exist different address types, which can be categorized
as follows:
● Local addresses versus global addresses: The local addresses are recognized
locally within the network as private addresses, while the global addresses are
used as public addresses which can be routed through the Internet.
● Temporary addresses versus persistent addresses: The temporary addresses are
assigned to network devices for a short period of time with dynamic host con-
figuration protocol, while the persistent addresses are allocated permanently to
specified nodes in the network.
3.1.2 Addressing mechanisms

In order to manage the IP addressing of devices in a network, it is necessary to have
different mechanisms to classify the IP addresses, to divide the network into different
subnetworks and also to group the network addresses of different subnets. Specifically,
the IP addresses can be managed using either one of the following mechanisms or a
combination of these approaches:
● Classful addressing: The IP addresses are arranged in different classes, i.e., Class
A, B, C, D, and E, with default network mask, namely, natural mask.
● Subnetting: An IP address hierarchy is created to divide a network into
subnetworks sharing the same subnet mask.
● Variable-length subnetting: Different subnet masks are employed for subnets of
different sizes.
● Supernetting: The IP addresses of different subnets in a network are aggregated
prior to advertise to the Internet.
This chapter is focused on discussing the above-addressed mechanisms for IPv4 in the
following sections. Their extension for IPv6 addressing can be deployed in a similar
approach.
3.2 IP version 4 (IPv4)

An IPv4 address has a length of 32 bits which are represented in the form of four
octets. Each octet has 8 bits,1 and these four octets are separated by a dot and thus
also known as dotted-quad format. For simplicity, an IP address is usually written in
dot-decimal format consisting of four decimal numbers, each of which is in the range
of 0–255, i.e., 00000000 to 11111111 in binary format, for example, 128.10.10.82
(cf. Figure 3.3).
The IPv4 addresses can be allocated either locally as private addresses or globally
with classful network architecture.
1
Note that the term “octet” is different from “byte” which is used as a storage unit of 8 bits.
32 bits
8 bits
IPv4 in binary 10000000 00001010 00001010 01010010
1 st Octet 2nd Octet 3rd Octet 4 th Octet
IPv4 in dot-decimal 128 10 10 82
Figure 3.3 An IPv4 address
Table 3.1 Private IP address range
IP address range Number of addresses Mask CIDR
10.0.0.0–10.255.255.255 224 255.0.0.0 /8

172.16.0.0–172.31.255.255 220 255.240.0.0 /12
192.168.0.0–192.168.255.255 216 255.255.0.0 /16
3.2.1 Private IP addresses

As an attempt to cope with the issue of address exhaustion, the IETF has reserved IPv4
address ranges for LANs within home and enterprise which are regarded as private
networks. There are three private IP address ranges as listed in Table 3.1. These IP
addresses are only valid inside the network and cannot be routed to the Internet. In
order to connect to the Internet, an IP network address translator gateway is required
to map between private and public address spaces.
3.2.2 Classful IP addresses

Initially proposed for the Internet since 1981, IP addresses have followed a classful
network-addressing architecture. Given IP addresses of 32-bit length, the address
spaces have been divided into five classes, namely, Classes A, B, C, D, and E (cf.
Figure 3.4). Specifically, Class A, B, and C networks have different sizes, while Class
D addresses are only for multicast and Class E addresses are reserved for future use.
The classification of IP addresses is only based on the leading bits in the first octet.
The bit patterns, address ranges, and natural masks of different IP classes are defined
in Table 3.2 where N and H denote eight network bits and eight host bits, respectively.
Written in decimal format, the class of an IP address can be simply identified
through its first number as follows:
● Class A: 0–127
● Class B: 128–191
● Class C: 192–223
8 bits 24 bits
Class A 0 Network Host
16 bits 16 bits
Class B 10 Network Host
24 bits 8 bits
Class C 110 Network Host
Class D 1110 (For multicast address)
Class E 11110 (For future use)
Figure 3.4 IP address classes
Table 3.2 IP address classes
Class Leading Bit Natural CIDR Number of Number of

bits pattern mask networks hosts per
network
A 0 N .H .H .H 255.0.0.0 /8 27 − 2 224 − 2
B 10 N .N .H .H 255.255.0.0 /16 214 216 − 2
C 110 N .N .N .H 255.255.255.0 /24 221 − 2 28 − 2
D 1110 Not defined
E 1111 Not defined
● Class D: 224–239
● Class E: 240–255
Note that two IP addresses are taken away from the number of addresses per net-
work in Table 3.2 due to the usage of a network address with all-zeros host bits and
a broadcast address with all-ones host bits. There are also two Class A network
addresses 0.0.0.0/8 and 127.0.0.0/8 and two Class C network addresses 192.0.0.0/24
and 223.255.255.0/24 which are reserved and cannot be assigned to any network.
Let nX , hX , NN ,X , NH ,X , X ∈ {A, B, C}, denote the number of bits for Class X
network, the number of host bits in Class X network, the number of Class X networks,
and the number of hosts per Class X network, respectively. As shown in Table 3.2,
nA = 8 bits, nB = 16 bits, and nC = 24 bits. The length of an IPv4 address is 32 bits,
so the number of host bits is hX = 32 − nX [bits].
Since the leading bits are fixed and there are some reserved IP addresses, Nn,X
and Nh,X can be computed as
⎧
⎪ n −1
⎨2 A − 2, if X = A,
NN ,X = 2nB −2 , if X = B, (3.2)
⎪
⎩ nC −3
2 − 2, if X = C,
NH ,X = 2hX − 2 = 232−nX − 2. (3.3)
The total number of hosts in Class X network, denoted by NT ,X , is thus given by
NT ,X = NN ,X NH ,X . (3.4)
Example 3.2. Host IP addresses in different network classes:

● A host with an IP address 130.18.12.50/16 belongs to a Class B network having
network address 130.18.0.0, broadcast address 130.18.255.255, and (216 − 2) IP
addresses available for hosts in the range 130.18.0.1–130.18.255.254.
● A host with an IP address 200.168.120.5/24 belongs to a Class C network having
network address 200.168.120.0, broadcast address 200.168.120.255, and 254 IP
addresses available for hosts in the range 200.168.120.1–200.168.120. 254.
3.3 Subnetting
In a small network with a limited number of colocated hosts, the usage of only a
specific network address is possible to support all hosts sharing a common natural
mask. However, within an enterprise or a campus where a network is divided into
smaller networks or subnetworks, subnetting is crucial to allocate the IP addresses to
hosts in different subnets.
For subnetting, some bits for host identifier are used to indicate subnet number
(cf. Figure 3.5). The hosts in the same subnet share a common subnet address and
32 bits
n bits s bits h bits
IP address Network Subnet Host
Network prefix Subnet Host number

number
Subnet mask (n+s) bits “1” h bits “0”
Figure 3.5 Subnetting

subnet mask. Let n, s, and h denote the number of bits for network, subnet, and host,
respectively, i.e., n + s + h = 32 bits. The number of subnets per network, i.e., NS
and the number of hosts per subnet, i.e., NHS , are computed by
NS = 2s , (3.5)
NHS = 2 − 2 = 2
h 32−n−s
− 2, (3.6)
respectively. Notice that, in (3.6), two IP addresses are taken away for subnet address
and broadcast address. The total number of hosts per network with subnetting, i.e.,
NH , is thus given by
NH = NS NHS = 2s+h − 2s+1 . (3.7)
Remark 3.1. It can be observed that the number of hosts with subnetting is less
than those without subnetting due to the usage of additional IP addresses for subnet
address and broadcast address in each subnet. Indeed, without subnetting, the total
number of hosts per network is given as in (3.3), i.e., NH ,no subnet = 2s+h − 2 > NH .
Example 3.3. Given a network address 180.16.0.0/16, i.e., n = 16 bits for network
prefix. If we borrow 3 bits for subnetting, i.e., s = 3 bits, we have a total of NS = 23 =
8 subnets, each of which has NHS = 213 − 2 hosts per subnet. All eight subnets have
the same subnet mask 255.255.224.0 with CIDR /19. The subnet address, broadcast
address, and host address range in each subnet can be assigned as in Table 3.3.
Table 3.3 Subnetting in Example 3.3
Subnet Subnet address Host addresses Broadcast address
1 180.16.0.0 180.16.0.1–180.16.31.254 180.16.31.255

2 180.16.32.0 180.16.32.1–180.16.63.254 180.16.63.255
3 180.16.64.0 180.16.64.1–180.16.95.254 180.16.95.255
4 180.16.96.0 180.16.96.1–180.16.127.254 180.16.127.255
5 180.16.128.0 180.16.128.1–180.16.159.254 180.16.159.255
6 180.16.160.0 180.16.160.1–180.16.191.254 180.16.191.255
7 180.16.192.0 180.16.192.1–180.16.223.254 180.16.223.255
8 180.16.224.0 180.16.224.1–180.16.255.254 180.16.255.255
Remark 3.2. In Example 3.3, the third octet of the subnet addresses is multiples of
32, which is 28−s with s = 3. Note that this is only applied for n = 16 and s ≤ 8.
Extended for the scenario when 8 < s ≤ 16, the fourth octet of the subnet addresses
would be multiples of 216−s , while the third octet is unchanged. A similar scenario
can be observed when n = 24 and s ≤ 8. In this case the fourth octet of the subnet
addresses would be multiples of 28−s .
A typical problem is to design a subnetting scheme for a network consisting of

K subnets and n bits for network prefix. The kth subnet, k = 1, 2, . . . , K, is required
to have NHk hosts. The following steps can be employed:
Step 1 Identify the subnet with a maximum number of hosts, i.e.,
NHS ,max = max{NH1 , NH2 , . . . , NHK }.
Step 2 Find the minimum number of host bits that can support the maximum number
of hosts using (3.6), i.e., solving

NHS ,max ≤ 2h − 2 or h ≥ log2 NHS ,max + 2 (3.8)

The minimum number of host bits is thus hmin = log2 NHS ,max + 2 , where x
denotes the least integer greater than or equal to x.
Step 3 Find the number of bits borrowed for subnetting, i.e., s = 32 − n − hmin .
Step 4 Determine subnet mask and design subnets.
Example 3.4. Given an IP address 198.16.10.0/24, design subnetting for laboratories
with the required number of hosts in each lab as in Table 3.4.
Table 3.4 Laboratories in Example 3.4
Lab Users
A 60
B 30
C 25
D 20
In this example, Lab A has the maximum number of hosts with NHS ,max = 60.
From (3.8), the minimum number of hosts bits is hmin = 6 bits. The number of bits
borrowed for subnetting is s = 32 − 24 − 6 = 2 bits and thus can support up to four
subnets sharing the same subnet mask 255.255.255.192 with CIDR /26. Each subnet
can have maximum NHS = 62 hosts satisfying the requirements of the number of users
in all labs in Table 3.4. The subnet address, broadcast address, and host address range
in each subnet can be assigned as in Table 3.5.
Table 3.5 Subnetting in Example 3.4
Lab Subnet address Host addresses Broadcast address
A 198.16.10.0 198.16.10.1–198.16.10.62 198.16.10.63

B 198.16.10.64 198.16.10.65–198.16.10.126 198.16.10.127
C 198.16.10.128 198.16.10.129–198.16.10.190 198.16.10.191
D 198.16.10.192 198.16.10.193–198.16.10.254 198.16.10.255
3.4 Variable-length subnetting

The usage of fixed-length subnetting as in Section 3.3 can be applied to the network
having subnets of the same or most likely similar number of hosts. For the networks
having subnets of different sizes, e.g., Table 3.4, only one subnet mask taking into
account the subnet with maximum number of hosts is not efficient. Such inefficiency
can be easily noticed when subnets have considerably different sizes as in the following
example.
Example 3.5. A network has an IP address with a network prefix of n = 16 bits
and subnet identifier of s = 6 bits (cf. Figure 3.6). The network can have NS = 64
subnets, each of which can have NHS = 1022 hosts corresponding to h = 10 host bits.
However, for the subnets with much less than 1,022 hosts, e.g., only 100 hosts, this is
really a waste. In these small subnets, 7 bits are enough for the hosts, i.e., h = 7 bits.
In this case, the subnet identifier has a length of s = 9 bits, which can support up to
512 subnets.
32 bits
16 bits 6 bits
LAN 1 Network Subnet Host
Subnet mask 255 255 252 0
VLSM
16 bits 9 bits 7 bits
LAN 2 Network Subnet Host
Subnet mask 255 255 255 128
Figure 3.6 Variable-length subnet mask in Example 3.5
In order to deal with this practical issue, variable-length subnetting is crucial to

establish an address hierarchy within a network having subnets of different sizes. A
variable-length subnet mask (VSLM) should be used in which a longer mask is for
subnets with fewer hosts and a shorter mask is for subnets with more hosts.
As can be observed in Example 3.5, VLSM also implies subnetting of subnets
to reduce the address wastage. In other words, the VLSM is used to maximize the
efficiency in addressing. The subnetting should be designed to guarantee that host
addresses are not overlapped between subnets. Therefore, prior to design subnetting,
it is suggested to sort the subnets in descending order of the number of hosts. The
following steps can be employed:
Step 1 Sort subnets in descending order of the required number of hosts.
Step 2 Find the minimum number of host bits for each subnet that can support its
required number of hosts using (3.6).
Step 3 Find the number of bits borrowed for subnetting in each subnet with respect
to the number of host bits determined in Step 2.
Step 4 Determine subnet mask for each subnet, and design all subnets and host
address range.
Example 3.6. Given an IP address 199.10.10.0/24, design variable-length subnetting
for an office consisting of four rooms with the required number of hosts in each room
as in Table 3.6. Following the steps for the VLSM design, the subnetting for these four
rooms can be sketched out as in Table 3.7 where for simplicity the network part in the
host address range, i.e., 199.10.10, is eliminated. The subnet mask is also not shown,
which can be deduced from the CIDR.
Table 3.6 Rooms in Example 3.6
Room Users
A 12
B 52
C 20
D 5
Table 3.7 Variable-length subnetting in Example 3.6
Room Host Subnet CIDR Subnet address Host addresses Broadcast address
bits bits
B 6 2 /26 199.10.10.0 0.1–0.62 199.10.10.63

C 5 3 /27 199.10.10.64 0.65–0.94 199.10.10.95
A 4 4 /28 199.10.10.96 0.97–0.110 199.10.10.111
D 3 5 /29 199.10.10.112 0.113–0.118 199.10.10.119
3.5 Supernetting
While subnetting is employed to divide a network into subnetworks, supernetting is the
process of combining subnetworks into a large network or supernetwork. Supernetting
is viewed as network address aggregation. The requirement of supernetting is further
illustrated in Example 3.7.
Example 3.7. Assume that there are 12,000 computers in an enterprise to be net-
worked. A Class B network can be used to support up to NH ,B = 216 − 2 =65,534
hosts. However, this is a wastage of IP resources. Alternatively, 48 Class C network
addresses can be employed for NH ,C = 48(28 − 2) =12,192 hosts. In this case, for
routing, it is required to advertise 48 network addresses. Instead of advertising all of
these addresses, address aggregation or route summarization can be carried out to
form a supernet, and only the summarized address is advertised.
In order to perform the address aggregation, the subnet ID should be aggregated

with the following rules:
● The number of addresses in a CIDR block is always a power of two.
● There is no gap or hole in the address block sharing a common border.
● Two ends of the address block correspond to bit boundaries.
Example 3.8. A network consists of four subnets with the following/24 IP addresses
(cf. Figure 3.7):
32 bits
24 bits 8 bits
LAN 1 130.100.00000000. 00000000
LAN 2 130.100.00000001. 00000000
LAN 3 130.100.00000010. 00000000
LAN 4 130.100.00000011. 00000000
Aggregation 130.100.00000000. 00000000 / 22
Supernet mask 255.255.11111100 0000000
255 255 252 0
Figure 3.7 Supernetting in Example 3.8
● 130.100.0.0/24
● 130.100.1.0/24
● 130.100.2.0/24
● 130.100.3.0/24
Performing address aggregation, the above four IP addresses, can be grouped with
CIDR /22, and only 130.100.0.0/22 is advertised to represent all subnets.
3.6 IP version 6
IPv6 is the latest version of the IP developed by the IETF in 1995. The IPv6 addresses
are used as 128-bit identifiers for either interfaces or sets of interfaces. With such
extensive address space, the IPv6 is aimed to replace the existing IPv4.
An IPv6 address consists of 128 bits, i.e., 16 octets, which are represented in eight
groups of hexadecimal numbers. Each group has two octets, and these eight groups
are separated by colons. For example, 0182:0A1C:001F:0000:0000:0014:0000:0003.
For convenience in representation, the IPv6 can be abbreviated in a shorter format by

applying the following rules:
● Leading zeros in any group of 16 bits, i.e., two octets or four hexadecimal digits,
can be omitted.
● Consecutive 16-bit zeros can be replaced by a double colon, i.e., “::,” with a notice
that only one double colon is used in the shortened form to allow the determination
of the full address.
Example 3.9. Given an IPv6 address written in full (cf. Figure 3.8) as
0182 : 0A1C : 001F : 0000 : 0000 : 0014 : 0000 : 0003.
By applying the above rule, this address can be abbreviated by
182 : A1C : 1F :: 14 : 0 : 3.
The full address can be determined by filling the missing leading zeros and replacing
the double colon with zeros satisfying there are totally eight groups of hexadecimal
digits.
128 bits (16 octets)

16 bits
0182 0A1C 001F 0000 0000 0014 0000 0003
182:A1C:1F :: 14 : 0 : 3
Figure 3.8 IPv6 address in Example 3.9
There are basically three IPv6 types including the following:

● Unicast address: An unicast address is used as an identifier for a single interface,
and the packet is only sent to that interface.
● Anycast address: An anycast address is used as an identifier for a set of interfaces,
and the packets will be sent to the nearest interface in this group.
● Multicast address: A multicast address is used as an identifier for a set of
interfaces, and the packet is delivered to all interfaces.
An IPv6 packet typically has the format as in Table 3.8 where the header has a
fixed size of 320 bits or 40 octets consisting of
● Version (4 bits): v6, i.e., “0110.”
● Traffic class (8 bits): for differentiated services.
● Flow label (20 bits): for data types, e.g., speech, video, etc., with priorities.
● Payload length (16 bits): to identify the size of payload in bytes including possible
extension headers. By default, payload length is 64 kB.
● Next header (8 bits): to identify the extension headers if present.
Table 3.8 IPv6 packet format
Version (4 b) Traffic Class (8 b) Flow Label (20 b)

Payload length (16 b) Next header (8 b) Hop limit (8 b)
Source address (SA) (128 b)
Destination address (DA) (128 b)
Possible extension header
Payload
● Hop limit (8 bits): is set as a counter which is decremented by one at each inter-
mediary device, e.g., router/gateway and the packet is dropped when this counter
reaches zero.
● Source address (SA)/destination address (DA) (128 bits): is IPv6 address of the
source/destination node.
There are also extension headers which have the size of multiple of 8 bits to carry
additional information about the Internet layer; for instance, some have been defined
including hop-by-hop header, end-to-end header, routing header, fragment header,
authentication header, and privacy header.
3.7 Related works
references:
IP Addressing Postel in 1981 [51], Mockapetris in 1987 [52], Braden in 1989 [53],
Egevang and Francis in 1994 [54], Rekhter et al. in 1996 [55], Graziani and
Johnson in 2007 [28], Dye et al. in 2011 [3], and Cotton et al. in 2013 [56].
IPv4 Perkins in 1996 [57] and in 2010 [58], and Touch in 2013 [59].
Subnetting Mogul in 1984 [60] and Tsuchiya in 1991 [61].
VLSM Pummill and Manning in 1995 [62].
CIDR Rekhter and Li in 1993 [63], Rekhter in 1995 [64], and Fuller and Li in
2006 [65].
IPv6 Hinden and Haberman in 2005 [66], Hinden and Deering in 2006 [67],
Boucadair and Venaas in 2014 [68], and Deering and Hinden in 2017 [69].

[Q3.1] Describe two versions of IP addresses and differentiate between them.
[Q3.2] Explain how to determine the network address and reasoning of using
network mask.
[Q3.3] List all private IP address ranges and give an example of their usage.
[Q3.4] How many IPv4 address classes? Show their bit pattern/format with
corresponding natural mask.
[Q3.5] Describe the steps for subnetting a network, given the number of required
hosts in each subnet.
[Q3.6] Why do we need to consider variable-length subnetting and in which
scenarios?
[Q3.7] Compare fixed-length subnetting versus variable-length subnetting with
illustrations.
[Q3.8] Explain why supernetting is necessary in an enterprise via an example.
[Q3.9] What are the rules in address aggregation?
[Q3.10] What is the length of IPv6 addresses and how to represent them in
abbreviated format?
[Q3.11] How many IPv6 types? Describe and differentiate between them.
[Q3.12] Describe the format of an IPv6 packet.
3.9 Problems
[P3.1] A small organization has a Class C IP address for 6 networks each with 20
hosts. What is an appropriate subnet mask?
[P3.2] A university has 200 LANs with 100 hosts in each LAN. Design an
appropriate CIDR addressing scheme for supernetting.
[P3.3] Perform CIDR aggregation on the following /24 IP addresses:
128.56.24.0/24, 128.56.25.0/24, 128.56.26.0/24, 128.56.27.0/24.
[P3.4] What is the efficiency of IPv6 packets that carry 10 ms of 64 kbps voice?
[P3.5] Abbreviate the following IPv6 addresses:
● 0000:0000:0F56:6281:0AB0:6DB7:BB27:7323
● 0000:0000:0000:AD63:73BA:0000:87AB:0392
● 2918:00AD:0000:0000:0000:0000:0123:000A
[P3.6] Write the following IPv5 address in full:
● 0:10::78A:6D
● ::13B:4C:0:8
● ::1
[P3.7] Fill in Table 3.9 with Class C address and Table 3.10 with Class B address
Table 3.9 Subnetting of Class C address
Number of Number of Number of Total number Subnet mask/

subnetting bits subnets/networks hosts/subnet of hosts natural mask
0
1
2
3
4
5
6
Table 3.10 Subnetting of Class B address
Number of Number of Number of Total number Subnet mask/

subnetting bits subnets/networks hosts/subnet of hosts natural mask
0
2
4
6
8
10
12
14
[P3.8] Find the network address, broadcast address, and host address range given
a host address:
● 192.168.180.94/27
● 172.23.10.19 /21
● 147.252.238.20 255.255.240.0
[P3.9] Divide a network having an address 136.178.0.0/16 into 16 subnets. Show
the binary and dotted-decimal forms of each subnet, as well as the subnet
mask.
[P3.10] Given an IP address 196.168.1.0. Develop a subnetting scheme for the
network in Figure 3.9.
14 hosts
Switch1 PC1
BRANCH 1
20 hosts
Switch0
PC0
HQ
8 hosts 12 hosts
Switch2 BRANCH 2 BRANCH 3 Switch3

PC2 PC3
Network address: 196.168.1.0/24

[P3.11] Given an IP address 129.130.0.0/16. Develop a variable-length subnetting

scheme for the network in Figure 3.10.
LAN 1 LAN 4
Se1/0 Se1/: Fa0/0 Fa0/1 Fa0/2 Fa0
Fa0 Fa0/2 Fa0/1 Fa0/0
2811 Se1/0 11 2950-24
2950-24 PC-PT
PC-PT BRAN Se1/0 BRANCH 4 Switch4
Switch1 PC4
PC1
Network address: 129.130.0.0/16
LAN 1: 100 hosts

LAN 2: 200 hosts
LAN 3: 500 hosts
LAN 4: 1,000 hosts
Se1/1
LAN 2 se1/0 LAN 3
Se1/1 Se1/0=a0/0 Fa0/1 Fa0/2 Fa0
Fa0 Fa0/2 Fa0/1 Fa0/0
2950-24 2811 2811 2950-24 PC-PT
PC-PT
Switch2 BRANCH 2 BRANCH 3 Switch3 PC3
PC2
[P3.12] Given an IP address 150.100.0.0/16, the workgroups within an organi-

zation in Table 3.11 are required to be networked. Each workgroup is
connected to a router having two Ethernet LANs and two serial interfaces.
Design a variable-length subnetting scheme to meet the requirements of
the organization.
Table 3.11 List of workgroups in Problem 12
Department Number of workgroups Number of hosts per workgroup
Administration 1 400
Finance 1 40
Sale 1 150
Manufacturing 15 60
Training 2 20
[P3.13] An organization has 1,200 hosts required to be networked with only Class
C addresses, and the organization wants to advertise a single address. The
address range available is 194.20.1.0–194.20.100.0.
● Discuss the criteria to select an appropriate block of addresses for this
design.
● Design an addressing scheme to meet the network requirements.
[P3.14] A university has three sites, each of which has different departments with
different requirements as shown in Table 3.12. The university has been
allocated an IP address 132.128.0.0/16. Develop an IP addressing scheme
to meet the design requirements with the most efficient use of address
space.
Table 3.12 List of departments in Problem 14
Department Site Number of subnets Number of hosts per subnet
Administration A, B, C 2 per site 200

Finance A 1 30
Students A, B, C 20 per site 25
Academic A, B, C 7 per site 150–200
Services A 30 500
Headquarters A 1 30
Chapter 4
Network analysis, architecture, and design
The first three chapters have reviewed basic concepts of internetworking and network
fundamentals with different network models, routing protocols, and IP-addressing
mechanisms. Given these background knowledge and existing technologies, it is
absolutely necessary to understand how to model and design a network as well as
validate its performance in different settings or configurations. The characteristics
of systems and services also need to investigate with proper use of terminologies in
computer-networking context.
In this chapter, I will first introduce in Section 4.1 a brief overview of various
approaches and processes for network analysis, architecture, and design. Details of the
network analysis, architecture, and design processes will be sequentially presented in
Sections 4.2–4.4. As a typical network-design model, the hierarchical design approach
will be discussed in detail in Section 4.5 along with its layers in Section 4.6. Various
network-design approaches and their evolution will be summarized in Section 4.7.
Finally, Section 4.8 will provide further discussion on network management and
security.
4.1 An overview of network analysis, architecture, and design

Bandwidth has been well regarded as a key-performance metric in every network
design. A network designer at all times aims at achieving a high bandwidth. The
bandwidth is indeed of great significance, but it is not the only one component of
network resources. There are other metrics that need to be considered in a practical
network design such as latency, reliability, maintainability, and availability. Due to
the heterogeneity of the practical networks, there exist a number of constraints and
limitations, for instance, hardware configuration, network topology, transmission
media, capital expenditures (CAPEX), and operating expenses (OPEX).
Prior to performing network architecture modeling and design in a real system,
an initial analysis is crucial for drafting a preliminary network model. This analysis
process basically consists of requirement analysis, flow analysis, and/or risk analysis.
Based on the analyses, the relationships between functions within the network can
be set and modeled via a descriptive network architecture. Such network modeling
is helpful in determining an appropriate technology, equipment, and connectivity for
network-design process as well as performance evaluation.
Given:
• Historical data
• Observation data
• Survey data
• Requirements
• Goals
• Data collection and observation

• Requirement analysis
Network analysis
• Flow analysis
• Risk analysis
• Descriptive network architecture

• Interactions
Network architecture
• Trade-offs
• Constraints/Dependencies
• Technology selection
• Equipment selection
Network design
• Physical devices and connections
• Performance evaluation
Figure 4.1 Network analysis, architecture, and design process
The information flows in network analysis, architecture, and design should go

through the following steps (cf. Figure 4.1):
● Defining problem(s) to be addressed;
● Establishing and managing customer expectations;
● Monitoring the existing network(s), system(s), and environment;
● Analyzing the collected data;
● Developing a set of options to solve the developed problem(s);
● Optimizing options based on various trade-offs;
● Selecting one or more options;
● Planning the implementation;
● Designing network; and
● Evaluating network performance.
In the following sections, I will briefly introduce the key processes in accordance
with those described above.
Network analysis, architecture, and design 81
4.2 Network analysis process

In order to perform any network analysis, a network designer needs to have a back-
ground knowledge on networking and prerequisite skills. Specifically, it is required
to know network components from different network devices, such as switches, hubs,
and routers, to requirements and performance metrics for validation. Therefore, it is
imperative to study characteristics of network components and their behaviors under
various situations of network settings.
In the network-analysis process, the data fed in should consist of as much infor-
mation as possible. These information can be collected via paperwork by asking the
customer to fill in a survey form or through the observation and validation of the
network designer. This may consist of the following:
● State of the current network (if existing),
● Historical problems of the network (if any),
● Network goals from the customer, and
● Requirements from users, applications, devices, and network in general.
After collecting and analyzing all the above facts, as the output of the network analysis
process, the following information can be drawn:
● Descriptions for the requirements of the network in either a sorted list or a tabular
format,
● Demands of the traffic flows in different networks/subnetworks,
● Mappings of applications and devices to the network under design or development.
4.3 Network architecture process

Network architecture is a process which uses information from network analysis pro-
cess to develop a high-level and end-to-end structure for the network. Various network
functions are built during the network-architecture process, which consist of network
addressing, routing protocols, network management, performance evaluation, and
security control. The network architecture process hence plays a decisive role as a
critical requirement in every network design.
The network architecture process aims at developing a reference architecture
for the real network design. Various simulation software and tools, such as Packet
Tracer, Wireshark, NS2/NS3, etc., can be used to model the network architecture,
validate the network performance, and also to optimize the network settings subject to
various constraints. A well-developed reference network architecture should provide
the following:
● A description of the complete network considering all of its architectural network
components,
● A compilation of the relationship between network functions, and
● A description of their interactions, trade-offs, dependencies, and constraints
employed to arrive at the developed architecture.
4.4 Network design process

Network design is a process to provide physical details for the developed network
architecture. The network design process basically consists of the following tasks:
● Evaluate and choose appropriate technologies for each area/part of the network
under design.
● Develop strategies to connect these technologies across the network.
● Learn how to set design goals, such as minimizing CAPEX/OPEX or maximizing
network performance, as well as how to achieve these goals.
● Apply the trade-offs, dependencies, and constraints developed as part of the
network-architecture process in solving the optimization problems.
● Develop evaluation criteria for the proposed design according to various perfor-
mance metrics.
Two typical trade-offs in the network design are generally discussed for every
network design, which are as follows:
● Cost versus performance: It is expected to have a high-performance network

with high bandwidth, low latency, and high reliability. However, such enhanced
performance normally requires a higher cost in both CAPEX and OPEX. This
is indeed a challenging problem that every network designer has to deal with
to balance the cost and network performance when considering various network
constraints as well as requirements and affordability of customers.
● Simplicity versus functionality: Taking into account the practical network deploy-
ment, the simplicity, and adaptability of a network design should be substantial in
facilitating the network-design process, while providing diverse functionalities as
per requests. An equilibrium between the network simplicity and network func-
tionality should be determined or optimized subject to the proposed constraints
and/or dependencies of the network components and functions.
For ensuring the network connectivity, hierarchy and interconnectivity models

are normally deployed across the network components (cf. Figure 4.2). The network
hierarchy design depends on the degree of concentration of networks or traffic flows
at interconnection points as well as the number of tiers or layers of interconnection
points within the network. The network hierarchies provide a separation between
tiers of points for a structured network when the network grows in size with a large
number of devices, applications, and services. The hierarchical design is also helpful
in determining the size of the network, addressing mechanism, scaling of network
technologies, and service levels within the network. In addition to the hierarchy, inter-
connectivity is deployed to improve the network reliability with redundancy playing
the role as backup links.
Interconnectivity
Interconnectivity
Networkhierarchy
Hiearchy
Interconnectivity
Interconnectivity
Network
Figure 4.2 Network hierarchical design with interconnectivity
4.5 Network hierarchical design

As described in the previous section, hierarchical design has been employed in most
of the networks. Some benefits of a hierarchical design model can be listed as follows:
● Scalability: The hierarchy allows the network to grow but still be able to control
and manage. All functionalities are localized, and potentials problems can be
easily identified to resolve. PSTN is a typical example of a very large-scale
hierarchical network.
● Implementation: With a hierarchical design, a network designer can assign clear
functionality to each layer, which helps facilitate network implementation.
● Troubleshooting: Since the functions of each layer are well defined, the problems
in the network can be easily isolated for fault detection and correction. The net-
work can be even temporarily segmented to reduce the scope and spreading of
the problems.
● Predictability: With the functional layers, the behavior of a hierarchical network
is predictable. In a sense, this is helpful for capacity planning to deal with network
growth in future. Such design approach also facilitates network modeling with
anticipated performance.
● Compatibility: The well-organized underlying infrastructure with hierarchical
design enables the adaptation of diverse applications and protocols.
● Manageability: The above-listed benefits of the hierarchical design contribute to
a great manageability of the network. As defined in the ITU standard for network
management, five domains of FCAPS can be maintained, including Fault Man-
agement, Configuration Management, Accounting Management, Performance
Management, and Security Management.
Core layer
Distribution
Network hierarchy
layer
Access layer
Figure 4.3 Network hierarchical layers
4.6 Network hierarchical layers
Basically, a network hierarchical design consists of core layer, distribution layer, and
access layer (cf. Figure 4.3). In the following sections, these three layers will be
described in detail.
4.6.1 Core layer

Core layer is at the top of the hierarchical design model. The core layer provides fast
wide-area connections between geographically remote sites, for instance, between
CANs within an organization or a large area connecting WANs (cf. Sections 1.1 and
1.3). In the core layer, there are rarely any hosts, and the links, i.e., core links, are
usually point-to-point links.
The services provided at the core layer are typically leased from telecom service
providers, e.g., Frame Relay (cf. Section 1.3.3.6). The major function at this layer is
to transport data efficiently between remote sites by providing redundant paths and
load sharing with rapid convergence and high bandwidth efficiency.
4.6.2 Distribution layer

Distribution layer is referred to as the layer-distributing network services to multiple
LANs within a campus or an enterprise network environment (cf. Section 1.2). The
campus backbone is deployed at this layer which is typically based on Fast Ethernet.
Core
Figure 4.4 One-layer design
The distribution layer is generally implemented on large sites which require to

interconnect buildings. As an intermediary to the top core layer, this layer mainly
supports policy-based connectivity that allows to control access to services, path
metrics, and also advertisements throughout the network.
4.6.3 Access layer

Access layer generally indicates a LAN or a group of LANs employing either Ethernet
or Token Ring to enable user access to network services. The access layer consists of
servers and workstations of all kinds.
The hosts, i.e., user workstations, within a workgroup of common interest can
be grouped and connected to the backbone of the network at distribution layer. The
access layer not only supports logical segmentation but also allows broadcast traffic
between the workgroups.
A practical network may have different number of layers and may not have all the
above three layers. Examples of some hierarchical designs are shown in Figures 4.4
and 4.5 for one-layer and two-layer designs, respectively.
Although the hierarchical design is shown to have a number of advantages (cf.
Section 4.5), the flows are forced to go through hierarchy, which may affect the
performance. When the connection between the access layer and distribution layer is
broken or any intermediary nodes are faulty or malfunctioning, the communications
between workgroups in different LANs or WANs are also lost. Hence, it is required
to provide redundancy with interconnectivity between layers. Furthermore, content
delivery network can be considered to bypass the network hierarchy for enhanced
performance (cf. Figure 4.6).
Core
Figure 4.5 Two-layer design
4.7 Network design approaches and evolution
In summary, a systematic approach for network design consists of the following main
steps:
● Step 1: Define problem(s)
● Step 2: Analyze customer’s requirements
● Step 3: Analyze network environment
● Step 4: Analyze current systems/networks (if any)
● Step 5: Propose solutions for the developed problem(s)
● Step 6: Perform theoretical modeling and analysis (optional)
● Step 7: Model the network with simulation software/tools
● Step 8: Determine the optimal solution
● Step 9: Plan the implementation
● Step 10: Design the practical network or test bed
● Step 11: Evaluate the network performance.
In the above steps for network design and analysis, network services and service-
based networking should be taken into account along with various performance
metrics, such as throughput, latency, and reliability. Parameters in network configu-
ration, such as bandwidth, flow allocation, networking, and physical design, should
also be evaluated to realize their impacts on the network performance, and hence
characterizing services with concluding remarks.
Network hierarchy
Interconnectivity
CDN
Figure 4.6 Network hierarchy over content-delivery network (CDN)
Network design is in fact an “art” in which a network designer is required to have

a good knowledge on technologies, services, and protocols. Given these background
along with working experience, appropriate network technologies can be selected
to meet the practical network-design problem. There are, however, too many options
with a number of available technologies, interconnection strategies, routing protocols,
and topologies. The design problem can be viewed as an optimization problem where
different performance metrics are required to be either individually or jointly opti-
mized subject to various constraints in practical networks as well as QoS requirements
from the customers.
4.8 Network management and security

The network design steps in Section 4.7 can only be realized with a good network
management. Such a well-managed network however exists together with attackers
and eavesdroppers who are trying to overhear the data or even harm the communi-
cations between devices across the networks. The security is therefore required to be
maintained in any network together with the network management to defend against
cyber threats and data breaches. The network security is a broad field consisting of
a number of policies and mechanisms in the literature from various perspectives at
different layers in the OSI reference model (cf. Section 1.4). The study of the network
security is only briefly introduced in this book. For a thorough understanding, the
readers are suggested to refer to the relevant works provided in Section 4.9.
A network designer should build a set of diagrams which can be used to prepare
the environment for the network design. These diagrams may include network design
goals and criteria for technology evaluation which will be discussed in Sections 4.8.1
and 4.8.2, respectively. Such schematic and illustrative representations are really
helpful for the network implementation with equipment settings, cable deployment,
etc. and also facilitate the network management afterwards. In particular, when the
design is based on an existing network, a network transition plan should be carefully
developed taking into account other factors, such as compatibility, adaptability, and
extendability.
4.8.1 Network design goals

In a typical network, design goals should be established subject to requirements of
the customers and specification of the network environment built or collected in the
above initial steps. Some of the common design goals can be listed as follows:
● Network performance should be optimized over various metrics, such as max-

imize network throughput, minimize latency, maximize reliability, minimize
energy consumption, minimize outage, maximize coverage, etc.
● Network deployment and/or operation should be at the lowest cost, i.e., minimize
CAPEX and/or OPEX.
● Management and maintenance should be as simplest as possible in a flexible man-
ner, i.e., minimize complexity in terms of computation, algorithm, mechanism,
policies, etc.
● Network security should be borne in mind at the highest priority, i.e., minimize
threats/attacks, minimize secrecy outage, optimize authentication, etc.
● Network should be adaptable to not only current user needs but also possible
extension in future, i.e., maximize network availability in the long term with
backup and extension plans.
Depending on the critical requirements in different networks, the above design goals
should be prioritized as in a wish list where we set primary goal(s), secondary goal(s),
etc. In order to satisfy these goals, the network designer is also required to consider
various constraints in the network. These constraints could be limited resources or
facilities, and cost in terms of both CAPEX and OPEX.
4.8.2 Technology evaluation criteria

The next step in the network-management process is the technology management to
meet the design goals. This technology management is based on the criteria used to
evaluate the technology selected for the network design. Given a number of existing
technologies nowadays, the technology evaluation is a challenging process where its
starting point is the list of design goals. A combination of various technologies is
normally considered in the practical network design. In particular, their adaptability
is crucial in the network management to make sure that they can support dynamic
behavior in the network, such as reconfiguration of user groups, addresses, routing,
and location.
The technologies are generally characterized through their capabilities with func-
tions provided in the network, such as network throughput, data flow, broadcast,
reliability, backup, security, etc. Specifically, both wired and wireless media in the
network with either connection-oriented or connectionless communication protocols
need to be managed properly guaranteeing that the capacity and service plans are in
a good progress.
As per required in any network design, the functions and features of technology
should be adaptable to users and applications with strict performance requirements.
The lack of communication infrastructure due to limited resources and costs should
also be managed in a flexible way with appropriate technologies and communication
protocols. For instance, some questions that could be raised are as follows: (i) How
to manage the multiple access in a crowded network with the usage of wireless tech-
niques and protocols? and (ii) How to optimize the data flows in the network which
is normally covered with asymmetric flows?
Considering the wireless users, their mobility together with their resources is
required to be investigated and managed in the network plan. The requirements of
multiple data rates for different services are also vital in these days in the selection
of the communication medium type with respect to different operating distances.
Moreover, the scalability of the network depends on the network connectivity in which
hierarchy and interconnectivity models are normally deployed for easy handling of
the data flows. Various medium types and technologies employed in a LAN can be
referred to in Section 1.2.3 in Chapter 1.
4.8.3 Network-management design

Network management is crucial in any networks from small to large sizes. The network
management is not simply polling simple network management protocol (SNMP)
information across the network, but it also requires monitoring, storage, and eval-
uation of different metrics, such as the number of bytes or IP packets in/out of an
interface and the number of errors at an interface.
This process should be logically designed to manage the whole network, and
thus it is generally integrated into the network design. As described in the previous
Sections 4.8.1 and 4.8.2, the network design goals and criteria for technology evalu-
ation should be taken into account to define and characterize the management for a
network design. Then, based on this, we can plan for monitoring and configuring the
network, as well as for troubleshooting in the case of network faults. These network
management protocols require to be examined regularly to update the instrumentation
requirements.
In summary, the network management involves the following tasks:
● Event notification via monitoring: The event could be a problem or a failure of
any components in the network, e.g., hosts, switches, hubs, routers, cables, NICs,
etc. A threshold is set based on the characteristics of the individual element,
which can be measured either between elements or across the entire network. The
characteristics may include capacity, delay, availability, reliability, utilization,
etc. The monitoring allows us to notify the status of the events in the network for
real-time analysis.
● Performance evaluation via monitoring: The monitoring process is not only for
examining events but also for evaluating the performance based on the collected
data. In order to analyze the events in the real time, it is required to have short
polling intervals which can only be supported in the network with high-speed CPU,
memory, and sufficient storage capacity. The performance metrics are validated
with reference to the preset baselines for every component within the network. In
the performance evaluation, link and element characteristics should be specific
to the type of the element being managed. For the sake of visualization, the
processed data, either real-time or time-averaging data, can be displayed, and the
old data is also required to archive for backup (if necessary).
● Network configuration: For network configuration, a characteristic set should be
generated for all elements in the network. Specifically, for each element, a table
of configuration parameters is generated along with techniques or methods which
can be employed to adjust these parameters. It is therefore required to understand
the advantages and disadvantages of these methods.
● Network troubleshooting: It is crucial to understand the effects of problems or
faults in the network and how to fix them. The troubleshooting task is a com-
plicated process requiring a thorough understanding of the characteristics and
parameters of all elements in the network. Through troubleshooting, problems
can be identified, noticed, isolated, and resolved.
● Planning: The above monitoring, evaluation, configuration, and troubleshoot-
ing tasks are all helpful for planning of the network management. For instance,
troubleshooting task helps us plan ahead of potential issues in the network, while
monitoring and evaluation support the reconfiguration of the network in the case
of faults.
The network management is generally designed with manageable resources by
employing appropriate network-management protocols and instrumentation. In the
context of computer networks, the instrumentation is the set of facilities that allows
us to access characteristics and configuration parameters of elements in the network.
The network-management protocols are used to couple this instrumentation with the
monitoring, processing, visualization, and storage. Specifically, SNMP is widely
used for the network management. It can provide facilities for not only collecting
but also configuring parameters from the network elements. The thresholds can be
set in the form of traps for automatic notices. In the SNMP, there is a management
information base which is a collection of parameters accessible by the SNMP for
organizing elements in the network. The SNMP also supports remote monitoring
allowing the MIP to provide information about the elements as well as the entire
network.
Similar to network design, network management also requires an architecture.
A network designer should consider the data flows for the information management.
The monitoring can be realized via either in-band or out-of-band monitoring with
the employment of centralized and/or distributed monitoring nodes. Due to these
additional tasks, capacity and delay requirements for the network-management need
to be taken into account. For instance, a monitoring node is deployed in every subnet
of a LAN. In this case, the traffic for network management data should be designed
so as not to overload the network.
4.9 Related works

references:
Network Analysis, Architecture, and Design Minoli in 1993 [70], Partridge in

1994 [71], Spohn et al. in 2002 [1], Pioro and Medhi in 2004 [72], Stallings
in 2006 [2], and McCabe in 2007 [73].
Network Hierarchy Raza and Turner in 1999 [74], Ravasz and Barabasi in 2003
[75], and Cisco Networking Academy in 2014 [76].
Network Management McCabe in 2007 [73], Oppenheimer in 2010 [77], and White
et al. in 2010 [78].
Network Security Denning in 1982 [79], Stallings in 2002 [80], Kent and Seo in
2005 [81], Jiang et al. in 2005 [82], Shiu et al. in 2011 [83], Liao et al. in
2013 [84], Pescatore in 2014 [85], Deka et al. in 2015 [86], Le et al. in 2015,
2016 and 2017 [87–90], Vien et al. in 2016 and 2017 [91–93], and Karim and
Vien in 2017 [94].

[Q4.1] Describe network-analysis process. What are the required inputs and
expected outputs of this process?
[Q4.2] What is the aim of network architecture process?
[Q4.3] How to validate a network architecture?
[Q4.4] What are key tasks in network design process?
[Q4.5] Explain two typical trade-offs in network design. Provide example(s) in
telecommunication systems for each of them.
[Q4.6] Describe three main layers in a network hierarchy.
[Q4.7] Give an example of employing interconnection within the network hier-

archy and demonstrate how these interconnectivity links can improve the
network reliability.
[Q4.8] With a flowchart, describe steps in a network design.
[Q4.9] Discuss some common network design goals with examples.
[Q4.10] What are the criteria in the technology evaluation?
[Q4.11] Describe tasks in network management. From your point of view, what is
the most challenging task in the network-management process?
4.11 Problems
[P4.1] Select any simulation software and do the research on how to develop a ref-
erence model using that software. Discuss its advantages and disadvantages.
[P4.2] Network hierarchical design is stated to allow us dealing with network growth
in the future with its predictability. Elaborate this further via an example to
show its extendability.
[P4.3] Consider a university network having two campuses, each of which has
ten LANs. Develop a network hierarchical model for this network with the
employment of interconnectivity links.
[P4.4] Do the research on techniques in content delivery network (CDN) and show
how they can be integrated into the traditional network hierarchy.
[P4.5] In the network design, there is a step requiring us to determine the optimal
solution of the network model, i.e., Step 8. Discuss potential approaches
and/or available techniques to solve this optimal problem.
Chapter 5
Network-requirement analysis
Network-requirement analysis is a must-have process in every network design.

Although this step seems to be simple, it plays an important role in the future network
modeling and implementation. First, it is vital to understand the existing systems
and network services. The requirements from the customer also need to be collected,
based on which requirement analysis model and framework can be built.
In this chapter, I will first introduce various network services and requirements in
Section 5.1. Service characteristics and levels will be described in Section 5.2 along
with discussion on requirements of service performance. Section 5.3 will provide
detailed steps of requirement analysis process where user, application, host, and
network requirements will be sequentially presented in Sections 5.4–5.7. A typical
requirement analysis model will be developed in Section 5.8. This chapter further
provides analyses of reliability, maintainability, and availability (RMA) of the network
and its components in Section 5.9.
5.1 Network services and requirements
Network-design process is not simply modeling and implementation. It requires a

careful preparation taking into account several factors in a whole. A network designer
should set a clear design goal according to practical circumstances. It could be either
maximizing network performance or minimizing the cost, i.e., CAPEX or OPEX,
or balancing both of them. The network architecture/topology and required func-
tionalities also need to be considered in advance of the modeling and deployment.
Identifying the services and performance levels in accordance with the requirements
and design goals is crucial prior to deciding any network technologies.
With respect to network services and service-based networking, levels of various
performance metrics, such as latency, bandwidth, reliability, availability, and security,
should be taken into account. Starting from the first generation in the 1960s with
only basic connectivity to the recently developed fourth generation, the service-based
networking is expected to provide self-configuration and administration.
A system is defined as a set of components that are designed to work together
as a whole. These components allow connectivity, communications, and services
to users in the network. A system in general consists of four generic components
(cf. Figure 5.1) which are as follows:
User User
Application Application
Network
Host Host
Figure 5.1 System components
● Users: Developers and customers both are users in the system.

● Applications: Applications can be either generic to the entire network or specific
to a particular user per request.
● Hosts: Hosts in a system are either computing devices, servers, or specialized
equipment which are connected together.
● Networks: Networks consist of all component interfaces within the system to
provide various services.
In IP networks, services and their descriptions are developed by the Internet Engi-
neering Task Force (IETF). The IETF standard reports and specifications also define
levels of performance, e.g., capacity, delay, availability, and reliability, and functions
supported in the system, e.g., management, scheduling, security, accounting, and
billing.
5.2 Service characteristics

The performance levels and functions offered by either hosts, applications, or net-
works also reflect service characteristics that can be offered to meet the requests.
The service characteristics, which will be described in detail in the following section,
are of paramount significance in configuring network elements, e.g., hubs, switches,
Network-requirement analysis 95
routers, OSs, etc. Indeed, they are helpful in not only network configuration but also
monitoring and validation of various services in the system.
In order to facilitate the description of service levels in the network, service
characteristics are grouped together. A service level can be then configured, measured,
and verified instead of a number of service characteristics. The service levels are
accordingly beneficial in network accounting and billing. These service levels can be
described by the following factors:
● Committed Information Rates (CIRs): In a network where virtual circuit is
employed for a specified service, CIR is defined as a guaranteed data rate, and the
bandwidth (cf. Definition 1.2 in Chapter 1) should not fall below this committed
figure.
● Classes of Service (CoS): CoS is literally a priority value that can be employed to
differentiate traffics of different kinds.
● Types of Service (ToS): ToS is employed to designate the priority of datagram
allowing to determine how the datagram should be routed for different services.
● Quality of Service (QoS): QoS is specified to guarantee a certain level of perfor-
mance to a data flow in the network targeting either low-latency, high-throughput,
or high-reliability services.
Service requests in general are classified as either best effort service or specified
service with deterministic functions. In addition, service levels need to be defined
so as to meet these requests. There are a variety of typical service performance
requirements with respect to different metrics as follows:
● Reliability: is a measure of the accuracy of information delivery in the system.
This can be represented by either bit error rate (BER) (cf. Definition 5.1) or packet
error rate (PER) (cf. Definition 5.2). The reliability can also be reflected through
the completion time to notify the accuracy of the information. If the completion
time of a data transmission takes longer than that is specified for every ToS, then
the data is regarded as lost or corrupted.
● Capacity: is a measure of the amount of information that can be reliably trans-
ferred over the communication channels. In computing, the bandwidth is regarded
as the theoretical capacity of one or more elements in the system (cf. Definition
1.1). The capacity is also referred to as the maximum bit rate which is measured
in bits per second (bps) (cf. Theorems 5.1 and 5.2 for the well-known Nyquist’s
and Shannon’s theorem to calculate the channel capacity of noiseless and noisy
channels, respectively). In the real system, throughput is normally measured as
the achievable capacity of the system or its component which is lower or much
lower than the theoretical capacity.
● Latency: is a measure of time differences in transmission of information between
nodes across the system incorporating application-processing time and task-
completion time. Note that the latency is different from network delay which
specifies only the duration for data to travel across the network. The network
delay generally consists of processing delay of the packet headers at the end
node, queuing delay of the packets in the waiting queues, transmission delay
for pushing packets onto the communication link, and propagation delay for the
signals to travel from the source node to the destination node.
Definition 5.1. BER is defined as the number of bit errors divided by the total number
of transmission bits during a time interval. These bit errors can be caused be noise,
interference, and distortion in digital transmission. The BER is normally represented
in percentage with no unit.
Definition 5.2. PER is defined as the number of incorrectly received packets over
the total number of packets transmitted over the media. The PER is expressed in
percentage, and the packet is incorrectly received when there is at least a bit error.
From the above definition of the BER and PER, their relationship can be deduced
as follows:
Lemma 5.1. Consider a data transmission with data packets having length of N bits.
Let Pe and PN denote the BER and PER, respectively. The PER can be determined
through the BER as
PN = 1 − (1 − Pe )N . (5.1)
Proof. As defined in Definition 5.2, a packet is incorrectly received when there is at

least a bit error. This means that a packet is only correct when all N bits in that packet
are correct, i.e.,
1 − PN = (1 − Pe )N . (5.2)
The lemma is proved.
Corollary 5.1. For large packet size with low BER, i.e., N → ∞, Pe → 0, (5.1) can
be approximated as
PN ≈ NPe . (5.3)
Proof. When N → ∞ and Pe → 0, it can be approximated that
(1 − Pe )N ≈ 1 − NPe . (5.4)
Substituting (5.4) into (5.1), we have (5.3). The corollary is proved.
An illustration of PER is shown in Figure 5.2 where BER Pe = 0.001.
Theorem 5.1 (Nyquist’s Theorem). Consider a bandwidth-limited signal consisting

of V discrete signal states transmitted with a bandwidth of B [Hz] over a noiseless
channel. The channel capacity, denoted by C, can be computed by Nyquist’s theorem
as follows:
C = 2B log2 V . (5.5)
1
Exact
0.99 Approximated
0.98
0.97
0.96
PN
0.95
0.94
0.93
0.92
0.91
0.9
0 10 20 30 40 50 60 70 80 90 100
N
Figure 5.2 Packet error probability (PER) versus the number of bits
A special case of binary data without modulation, i.e., V = 2, the channel

capacity can be simply given by
C = 2B. (5.6)
Example 5.1. A noiseless 10 kHz channel can transmit binary data at a maximum
rate of 20 kbps.
Definition 5.3. The signal-to-noise ratio (SNR) γ is defined as the ratio of sig-
nal power to noise power. The SNR is a dimensionless quantity, which can also be
expressed in logarithmic decibel scale as
γdB = 10 log10 γ . (5.7)
Theorem 5.2 (Shannon’s Theorem). Over a noisy channel having bandwidth of B

[Hz] and SNR of γ (cf. Definition 5.3), the channel capacity C, in bps, is determined
by Shannon’s theorem as follows:
C = B log2 (1 + γ ). (5.8)
Example 5.2. A noisy 10 kHz channel link having SNR of 10 dB can transmit binary
data at a rate up to 34.59 kbps, while a link with 0 dB SNR, i.e., signal power is equal
to noise power, results in a lower capacity of 10 kbps (cf. Figure 5.3).
120
100
80
C [Kbps]
60
40
20
B = 10 kHz
B = 20 kHz
0
0 2 4 6 8 10 12 14 16 18 20
SNR [dB]
Figure 5.3 Channel capacity versus signal-to-noise ratio (SNR)
5.3 Requirement-analysis process

Prior to performing the analysis, it is critical to understand network environment.
Specifically, system requirements and characteristics need to be first identified and
gathered. Then, services and performance thresholds can be determined.
The knowledge of network environment helps the network designer to have a
better understanding as well as to anticipate the behavior and characteristics of the
network under development. Furthermore, payoffs should be taken into account in
the requirement analysis.
In fact, there exist a variety of objectives and options for selecting the network
technologies. Either stand-alone or combined approaches can be considered, where
the former requires us to solve an optimization problem for an optimal solution,
while the latter needs to find the adaptation of various strategies in the network. The
compatibility of network components is also necessary to be investigated with respect
to various user and application requirements. The network designer therefore should
have a clear understanding of the location and mechanism to offer services in the
network.
Requirement-analysis process basically comprises user, application, host, and
network-requirement analyses following the structure of a generic system between
two end nodes (cf. Figure 5.4). In the following sections, each of these tasks will be
described in detail.
User requirements
Application requirements
Host requirements
Mainframe
Network requirements
Figure 5.4 Requirement-analysis process between two end nodes
5.4 User-requirement analysis

Targeting at user requirements, a system should be adapted to various user environ-
ment to offer high service quality. The system is commonly desired to provide high
data bandwidth with reliable information transfer and storage. Some general user
requirements can be listed as follows:
● Timeliness of data,
● Interactivity between users,
● Quality of supplied services,
● Adaptability in different networks,
● Security of data communications and storage from attacker, and
● Affordability of networking.
Apart from user expectation for high QoS, a network designer also needs to have
knowledge on
● number of users in the network,
● location and distribution/density of users,
● expected growth of the network, and
● redundancy for enhanced reliability or data backup.
5.5 Application-requirement analysis

As a second component in the generic system, application-requirement analysis is
related to analyzing the performance of services that can be supported within the
network. The performance metrics may include the following:
● Latency: is used to validate the interactivity and timeliness.
● Reliability: is a measure of the quality, adaptability, and security of communica-
tions and services.
● Capacity: reflects the affordability of the system, the number and density of users,
and also predicts the network growth with possible redundancy.
Given these three specific application requirements, there are three correspond-
ing application types as follows:
● Real-time applications: The real-time applications are the applications that require
to have a strict timing relationship between source node and destination node
considering the duration of both data transfer and processing. The delay require-
ment is specified for each application to enable the interactivity in real time. The
speech and non-buffered videos are typical examples of applications requiring
real-time services. On the other hand, non-real-time applications are referred to
as the applications with a loose delay requirement, such as web browser and email
applications.
● Mission-critical applications: In mission-critical applications, the reliability mea-
sure is taken into account. With deterministic and/or guaranteed reliability, the
mission-critical applications aim to provide accurate delivery of data. In fact,
the loss of reliability means the information is undetectable or unrecoverable,
which may cause the loss of revenues and even loss of life due to corrupted
sensitive data. In these applications, best effort delivery is employed given the
constraints/requirements of the reliability.
● Controlled-rate application: In most applications, a high bandwidth is normally
expected for the best QoS. This requirement is critical in controlled-rate applica-
tions where the different applications and services may require different rate. For
instance, voice, video, teleconference/teleservice, and file transfer have their own
specified requirements of data rate. In these controlled-rate applications, simi-
lar to the mission-critical application, the data delivery follows the best effort
approach subject to the constraint on the required bandwidth.
In order to analyze the application requirements, the applications should be first

categorized in different application groups as defined above. The application group
allows us to easily determine the general performance characteristics. Some examples
of application groups are as follows:
● Visualization applications, e.g., weather monitoring/forecast and medicine man-

agement, require high performance in terms of data rate and latency. These
applications belong to real-time and/or controlled-rate applications.
● Telemetry control applications, e.g., aircraft control, require high-reliability
and low-latency data communications. These applications are thus regarded as
mission-critical and/or real-time applications.
● Cloud computing applications, e.g., parallel computing with cluster/distributed
cloud-based servers, require interactivity with low-latency data communica-
tion and computation. These applications are normally classified as real-time
applications.
● Web applications, e.g., Outlook Web based mail services and web-based inventory
check, are delay sensitive by some means but do not require high performance.
These applications are thus tolerant of bandwidth and/or reliability.
● Data transport applications, e.g., fire transfer with FTP, require high reliability
but delay- and bandwidth-tolerant. These applications are listed as mission-critical
applications depending on the importance of data to be delivered.
● Teleservice applications, e.g., teleconference and webinars, provide multicast
services to a number of users. They are expected to provide high capacity and low
latency, while still maintaining a reliable communication. Taking the capacity and
latency into account, these applications are grouped into real-time and controlled-
rate applications.
● Operation, administration, and maintenance applications in general require high
reliability, while the capacity and latency are optional. These applications are
grouped as mission-critical applications.
The applications after being categorized with specified requirements will be

connected to the next component of the system, i.e., host requirement analysis, via an
interface providing detailed information about group, type, location, and performance
characteristics of all applications in the system.
5.6 Host-requirement analysis

Host-requirement analysis is the third step in a generic system (cf. Figure 5.4).
Generally, a system consists of three types of hosts as follows:
● Generic computing devices: e.g., PCs, laptops, workstations, etc.

● Servers: e.g., web server, application server, database server, mail server, etc.
● Specialized equipment: e.g., sensors, supercomputers, parallel computing
systems, etc.
Each kind of hosts is characterized by either some or all of the following

performance measures:
● Storage performance
● Processor performance
● Memory performance
● Capacity performance
● OS performance
● Device driver performance.
In addition to the performance criteria, the location information of each equipment

is critical in the network design and planning.
5.7 Network-requirement analysis

Network-requirement analysis, though the last step in a generic system, plays an
important role in the system. This process considers the system as a whole linking all
requirements from the previous steps. In addition, the existing network is required to
be validated for its adaptability and interoperability. Specifically, the network size,
services, location, and performance characteristics all need to be taken into account
in the network-requirement analysis. A network interface of the system therefore
consists of the following components:
● Network size and optional scaling;

● Location and/or relocation information;
● Network and support services;
● Adaptability and interoperability; and
● Network performance in terms of latency, reliability, and capacity.
Some basic functional requirements to be planned at the network level can be listed
as follows:
● Management requirements: How to configure network given existing network

settings and user requirements as constraints? Then, how to monitor network
events and evaluate network performance? Lastly, how to trace and correct faults,
a.k.a. troubleshooting, in the running network?
● Security requirements: It is important to define security policies together with
other requirement analysis at the starting point to make them manipulable in the
future network design. This is also helpful for risk analysis.
● Financial requirements: Available funds need to be verified considering both the
CAPEX and OPEX.
● Enterprise requirements: Required services in the system, e.g., data type, voice,
video, printer, phone, fax, etc., need to be discussed in advance.
Customer
Gathering and listing

requirements
Survey data
Developing performance
metrics and thresholds
Data analysis
Characterizing network
behaviors and services
Figure 5.5 Requirement-analysis model
5.8 Requirement-analysis model
From the above analyses, guidelines for these requirements should be outlined for all
steps in a generic system. Basically, a requirement-analysis process model consists
of the following three steps (cf. Figure 5.5):
● Step 1 (gathering and listing requirements): This step requires us to determine

initial conditions of the network under design; for instance,
– Design type: new, modification, validation, etc.?
– Design scope: network size, coverage area, number of sites, etc.?
– Design goals: high bandwidth, low latency, high security, etc.?
– Other issues: related to politics, administration, finance, environment, etc.?
In order to collect all the above information, a network designer should work with
users in a constructive manner. These could be done by either asking the customer
fill in a survey form listing a number of relevant questions or organizing meetings
for Q&A.
● Step 2 (developing performance metrics and thresholds): As discussed in the
previous sections, there are three specific performance metrics to be considered,
which include the following:
– Reliability: is represented by error/loss rate, e.g., BER or PER. The avail-

ability, stability, and maintainability also need to be taken into account, for
which each needs to be set with a threshold. These will be described in detail
in Section 5.9.
– Capacity: is a measure of the maximum achievable data rate or peak data rate.
The minimum data rate with respect to data packet size and duration should
be defined as a threshold.
– Latency: is normally validated through end-to-end or round-trip delay where
application-processing time and task-completion time are negligible com-
pared to the duration of the data travel. The requirement of timeliness
and delay variants caused by human response and interaction can also be
considered.
● Step 3 (characterizing network behavior and services): Service levels should be
configured by using network-management protocols, such as Simple Network
Management Protocol at the application layer (cf. Sections 1.4.1.1 and 1.5). The
behavior also needs to be characterized based on usage patterns and/or history
data, for example, frequency and duration of application sessions and traffics.
5.9 Reliability, maintainability, and availability analyses
In order to analyze performance requirements of application, service, and network,

besides practical measurements, statistical metrics should be considered to provide
an anticipated performance given available numerical data. Specifically, RMA are
three typical performance measures which are defined as follows:
● Reliability: is a statistical indicator for the frequency of network failures or the
unscheduled outages of service. The reliability, in this sense, is a function of the
failure rate which can also be represented by mean time between failures (MTBF)
in a system. Here, the failure rate is defined as the number of failures during the
operation in a unit of time.
● Maintainability: is a statistical measurement of the time for restoring system from
faulty status to be fully operational. The maintainability is thus represented by
mean time to repair (MTTR).
● Availability: is defined as the ratio of time that a device is in operation over
its total time including both available and unavailable duration of the device.
The availability shows the relationship between the frequency of mission-critical
failures and the time to restore service. It depends on both MTBF and MTTR (cf.
Figure 5.6).
Definition 5.4. The reliability function, a.k.a. survival function, is defined as the
probability that the lifetime of a device exceed a given specified time. Let T denote
the device lifetime, R(t) denote the reliability of that device in a duration t, and Pr [A]
denote the probability of an event A. By definition, R(t) can be computed by
R(t) Pr [T > t]. (5.9)
TBF TBF
On
Off Time
TTR TTR TTR
Figure 5.6 Time between failure (TBF) and time to repair (TTR) of a system
Lemma 5.2. When the lifetime of a device follows an exponential distribution with a
failure rate λ, its reliability in a duration t can be computed by
R(t) = e−λt . (5.10)
Proof. According to (5.9), R(t) can be given by
R(t) = 1 − Pr [T ≤ t]
(a)
= 1 − FT (t)
(b)
= e−λt ,
where FT (t) denotes the cumulative distribution function (cdf) of T , (a) follows from
the definition of the cdf, i.e., FT (t) Pr [T ≤ t], and (b) follows due to the cdf of
an exponential random variable given by FT (t) = 1 − e−λt (details of the cdf and
exponential distribution will be discussed in Chapters 11 and 13).
Definition 5.5. MTBF of a device is defined as the average operating time of the
device between failures. Statistically, the MTBF can be expressed as its average
lifetime, i.e.,
MTBF = E[T ], (5.11)
where E[ · ] denotes the expectation operator and E[T ] is the expected value, a.k.a.
average value or mean, of T .
Lemma 5.3. When the lifetime of a device follows an exponential distribution with a
failure rate λ, its MTBF is given by
1
MTBF = . (5.12)
λ
Proof. The proof follows from the expected value of an exponential random variable,
i.e., E[T ] = 1/λ.
Remark 5.1. Equation (5.12) in Lemma 5.3 means the MTBF of the device is the
reciprocal of its failure rate. This is indeed true from Definition 5.5 of the MTBF.
Consider a system consisting of N devices. The devices can be connected in either

series or parallel or series–parallel combination. Let RS (t) and Rn (t), n = 1, 2, . . . , N ,
denote the reliability of the whole system and the nth device, respectively, in a duration
t. The lifetime of the devices is assumed to be independently but not identically
distributed. The system reliability in each case can be determined as follows:
Lemma 5.4. The reliability of a system having N devices connected in series (cf.
Figure 5.7) is given by

N
RS (t) = R1 (t)R2 (t) · · · RN (t) = Rn (t). (5.13)
n=1
Device 1 Device 2 Device 3 Device 4 Device (n–3) Device (n–2) Device (n–1) Device n
Figure 5.7 Devices connected in series
Proof. With serial connection as in a chain, a system only works when all connected
devices work properly. The system reliability therefore depends on all of its devices.
Let TS and Tn , n = 1, 2, . . . , N , denote the lifetime of the system and the nth device,
respectively. It can be noticed in a serial connection that the lifetime of the system is
the minimum lifetime of all devices, i.e.,
TS = min{T1 , T2 , . . . , TN } = min {Tn } (5.14)
n=1,2,...,N
From the definition of reliability (cf. (5.9) in Definition 5.4), the system reliability is
given by
RS Pr [TS > t]

(a)
= Pr [min{T1 , T2 , . . . , TN } > t] = Pr min {Tn } > t
n=1,2,...,N

N
= Pr [{T1 > t} ∩ {T2 > t} ∩ · · · ∩ {TN > t}] = Pr {Tn > t}
n=1
(b)
N
= Pr [T1 > t] Pr [T2 > t] · · · Pr [TN > t] = Pr [Tn > t],
n=1

N
= R1 (t)R2 (t) · · · RN (t) = Rn (t),
n=1
where (a) follows from (5.14) and (b) follows due to the fact that the lifetime of
devices is independently distributed. The lemma is proved.
Corollary 5.2. In a serially connected system, when the lifetime of devices follows
an exponential distribution, each of which has a failure rate λn , n = 1, 2, . . . , N , the
system reliability is given by
N
RS (t) = e− n=1 λn t . (5.15)
Proof. The proof follows from (5.10) in Lemma 5.2 and (5.13) in Lemma 5.4.
Lemma 5.5. The reliability of a system having N devices connected in parallel (cf.
Figure 5.8) is given by
RS (t) = 1 − (1 − R1 (t))(1 − R2 (t)) · · · (1 − RN (t))

N
= 1− (1 − Rn (t)). (5.16)
n=1
Device 1
Device 2
Device (n–1)
Device n
Figure 5.8 Devices connected in parallel
Proof. When N devices are connected in parallel over N branches, the system still
can work when one out of N devices works fine. This means that the lifetime of the
system is the maximum lifetime of all devices within the system, i.e.,
TS = max{T1 , T2 , . . . , TN } = max {Tn } (5.17)

n=1,2,...,N
Following the same approach as in the proof of Lemma 5.4, the system reliability can
be determined by
RS Pr [TS > t]

(a)
= Pr [max{T1 , T2 , . . . , TN } > t] = Prmax {Tn } > t
n=1,2,...,N
N

= Pr [{T1 > t} ∪ {T2 > t} ∪ · · · ∪ {TN > t}] = Pr {Tn > t}
n=1
N
(b)
= 1 − Pr [{T1 ≤ t} ∩ {T2 ≤ t} ∩ · · · ∩ {TN ≤ t}] = 1 − Pr {Tn ≤ t}
n=1
(c)
= 1 − (1 − Pr [T1 > t])(1 − Pr [T2 > t]) · · · (1 − Pr [TN > t] )
N
=1− (1 − Pr [Tn > t]),
n=1

N
= 1 − (1 − R1 (t))(1 − R2 (t)) · · · (1 − RN (t)) = 1 − (1 − Rn (t)),
n=1
where (a) follows from (5.17), (b) follows from Pr [A ∪ B] = 1 − Pr [Ā ∩ B̄ ] gen-
eralized for N events, and (c) follows from the complementary property Pr [X̄ ] =
1 − Pr [X ] and property of independently distributed random variables. Here X̄
denotes the complement of an event X . The lemma is proved.
Corollary 5.3. In a parallel-connected system, when the lifetime of the devices follows
an exponential distribution, each of which has a failure rate λn , n = 1, 2, . . . , N , the
system reliability is given by

N
RS (t) = 1 − (1 − eλn t ). (5.18)
n=1
Proof. The proof follows from (5.10) in Lemma 5.2 and (5.16) in Lemma 5.5.
Remark 5.2. Connecting devices in parallel is also to provide backup(s) in practical

systems. If the same device is deployed across all N branches, i.e., λn = λ0 ∀n =
1, 2, . . . , N , the system reliability is given by
RS (t) = 1 − (1 − e−λ0 t )N . (5.19)
Remark 5.3. The two connection cases in Lemmas 5.4 and 5.5 are also well known
in electronic circuits. In a practical system, the connection of devices is much more
complicated. They can be connected in series-parallel hybrid or combination. The
reliability of the system can be found by first dividing it into simple serial and parallel
branches. Then, compute the reliability of each branch and group them together for
the whole system.
Definition 5.6. The availability of a device is defined by the ratio of the average
working time of the device to the total time of the device in the system. Let A, Top ,
and Ttotal denote the availability, operation time, and total time, respectively. A is
computed by
E[Top ]
A (5.20)
E[Ttotal ]
Note that the average operation time of the device is in fact the MTBF, and the total time
consists of the operation time, i.e., MTBF, and the time when the device is unavailable
and needs to be repaired, i.e., MTTR. The availability A can be given by
MTBF
A= . (5.21)
MTBF + MTTR
Lemma 5.6. When both the lifetime and repair time of a device follow exponential
distributions with a failure rate λ and a repair rate δ, respectively, its availability is
given by
δ
A= . (5.22)
δ+λ
Proof. As shown in Lemma 5.3, the MTBF is the reciprocal of its failure rate λ.
Similarly, when the repair time of the device follows an exponential distribution with
a rate δ, the MTTR can be computed by
1
MTTR = . (5.23)
δ
Substituting (5.12) and (5.23) into 5.21, the lemma is proved.
Remark 5.4. Unavailability is sometimes considered as opposed to availability met-

ric. The unavailability is defined as the ratio of time that the device is unavailable,
i.e., being repaired, over the total time in the system. The unavailability, denoted by
U , is simply given by
MTTR
U 1−A =
MTBF + MTTR
λ
= . (5.24)
δ+λ
5.10 Related works

references:
System Components Bertsekas and Gallager in 1992 [95], Stallings in 2003 [25],
and White in 2010 [96].
Service Characteristics Braden et al. in 1994 [97], Huston in 2000 [98], Evans
and Filsfils in 2007 [99], Marchese in 2007 [100], Braun et al. in 2008 [101],
Papadimitriou in 2010 [102], and Baker et al. in 2010 [103].
Reliability Rubino in 1999 [104] and Proakis and Salehi in 2001 and 2007 [105,106].
Capacity Nyquist in 1928 [107], Shannon in 1948 and 1949 [108,109], Marks in
1991 [110], and Cover and Thomas in 2006 [111].
Latency Ramaswamy et al. in 2004 [112], Stallings in 2006 [2], and Kurose and
Ross in 2012 [113].
Requirement Analysis Jain in 1991 [114], Kotonya and Sommerville in 1998 [115],
Hay in 2003 [116], McCabe in 2007 [73], and Laplante in 2013 [117].
RMA Analyses Smith in 2011 [118], Ayers in 2012 [119], Elsayed in 2012 [120],
and Lienig and Bruemmer in 2017 [121].

[Q5.1] List generic components of a system. Differentiate between hosts and users.
[Q5.2] How to describe service levels and based on what factors?
[Q5.3] Describe typical performance metrics used to validate services in a system.
[Q5.4] Differentiate between latency and network delay in a communication
system.
[Q5.5] What are bit error rate (BER) and packet error rate (PER)?
[Q5.6] Derive the relationship between PER and BER.
[Q5.7] Describe Nyquist’s theorem for noiseless channel. Show the impact of the
number of discrete signal states on the channel capacity.
[Q5.8] Describe Shannon’s theorem for noisy channel. Discuss the impact of
channel bandwidth and signal-to-noise ratio (SNR) on the channel capacity.
[Q5.9] Explain the requirement-analysis process via a flowchart.
[Q5.10] Differentiate between real-time applications, mission-critical applications,
and controlled-rate applications. Provide an example for each kind of
application.
[Q5.11] In the host-requirement analysis, discuss how to characterize the perfor-
mance of each kind of hosts. Choose a specific equipment and list all of
its performance measures.
[Q5.12] What are functional requirements at the network level?
[Q5.13] Describe steps in the requirement-analysis model. From your opinion,
which step is principal and time-consuming?
[Q5.14] What are reliability, maintainability, and availability analyses?
[Q5.15] Illustrate the difference between mean time between failures (MTBF) and
mean time to repair (MTTR).
[Q5.16] What is the relationship between the device lifetime and its failure rate with
the assumption that the lifetime follows an exponential distribution?
[Q5.17] Derive the reliability of a system consisting of N devices in two following
cases:
● N devices are connected in series.
● N devices are connected in parallel.
[Q5.18] How to determine the availability of a device? Show its relationship with
the MTBF and MTTR.
5.12 Problems
[P5.1] Consider a noiseless 20 kHz channel.
● Find the number of discrete signal states required to achieve a channel
capacity of 64 kbps.
● Given a fixed channel bandwidth, how many states are required to
achieve a double channel capacity in general?
[P5.2] In Theorem 5.2, the signal-to-noise ratio (SNR) γ is given by
S
γ = ,
N0 B
where S [W] is signal power, N0 [W/Hz] is noise power spectral density, and
B [Hz] is channel bandwidth.
● Rewrite (5.8).
● Derive the channel capacity for a very low SNR regime when B goes to
infinity.
[P5.3] Over a noisy channel having bandwidth B = 10 MHz and S/N0 = 105
● Find the channel capacity.
● Find the maximum channel capacity that can be achieved over broad-
band given fixed S/N0 .
[P5.4] The mean time between failures (MTBF) for a newly installed component is
given as 2,500 h.
● What is the reliability over a week, assuming 8 h/day and 5 days/week
working?
● How would the reliability change for the component if a backup is used?
● Calculate the overall reliability of three identical components connected
in series without backup.
● Calculate the overall reliability of the above three-component system
assuming that each has a backup individually.
● Assume the mean time to repair (MTTR) of a component is 0.5 h.
– What is the long-term availability of this component?
– What is the availability for the given time interval?
[P5.5] A system of components is connected as in Figure 5.9. The MTBF and MTTR
of each type of component are given in Table 5.1.
X Y Z
Y Z

Table 5.1 Table for Problem 5
Component MTBF (h) MTTR (h)
X 5,000 6
Y 80,000 12
Z 150,000 48
A B C E F G I
B C E F H J
(a) (b) (c)
● Find the long-term availability and the availability of each component

over time.
● Find the reliability of each component and the overall system reliability
over time.
● Draw a block diagram if Y and Z are individually backup. Find the new
overall system reliability and compare with the reliability of the original
setting.
● Find the best backup strategy to improve the system reliability given
three above components. Draw a block diagram to depict your strategy.
Calculate the new system reliability.
[P5.6] Prove the best backup strategy for N components serially connected in a
system is when each component is individually backed up.
[P5.7] Given a system consisting of three subsystems connected in series as in
Figure 5.10 where the components in each subsystem are grouped by dotted
lines. Let RX (t) denote the reliability of component X at a specific time t.
● Find the expression for each of the following reliabilities:
– Reliability of the first subsystem R1 (t) in terms of RA (t), RB (t), and
RC (t).
– Reliability of the second subsystem R2 (t) in terms of RE (t) and
RF (t).
– Reliability of the third subsystem R3 (t) in terms of RG (t), RH (t),
RI (t), and RJ (t).
– Reliability of the overall system R0 (t) in terms of R1 (t), R2 (t), and
R3 (t).
A B C E F G I
A B D E F H I
(a) (b) (c)
●Redraw the diagram using only the components given above showing
the layout to give the highest reliability. Find the formula of R1 (t), R2 (t),
R3 (t), and R0 (t) for the new configuration.
[P5.8] A system is set as in Figure 5.11 where the components in each subsystem
are grouped by dotted lines. Denote the reliability of component X at a
specific time t by RX (t).
● Find the reliability of each subsystem and overall system.
● Redraw the diagram using only the components given above showing
the layout to give the highest reliability. Find the formula of reliability
for the new configuration.
Chapter 6
Network flow analysis
Given established requirements with specifications of user, application, host, and

network via the requirement analysis step as discussed in Chapter 5, this chapter will
analyze these requirements further to provide an end-to-end perspective based on the
characteristics of elements within the network. This process is known as flow analysis.
The flow in the context of computer or mobile networks is understood as “data” flow,
while it may have different meanings adaptable to different models and environment.
With the aim of supporting a number of services with different requirements and
specifications, the flow analysis is a vital process in every network planning and
design.
In this chapter, I will first introduce the concepts of flow, different flow types,
and an overview of flow analysis process in Section 6.1. Basic components of a flow
will be briefly shown in Section 6.2. Section 6.3 will outline various flow models in
detail. Flow boundaries, distribution, and specification will be sequentially described
in Sections 6.4–6.6. Finally, this chapter will be closed by sketching guidelines and
procedures for network flow analysis.
6.1 Flows
A flow is defined as a set of application and protocol information having common
attributes, such as source and destination addresses, information type, routing pro-
tocol, etc. During a single session of a running application, there always exists a
flow between two end nodes, which is associated with the end-to-end communication
between a source (or sender) and a destination (or receiver). The flows therefore can
be examined through the communications on the basis of links within the network,
which can be then adapted for the entire network.
The network design is generally aimed for either capacity planning or service
planning. The flows are studied in the context of these two network planning schemes,
which are categorized as follows:
● Capacity planning: In this planning scheme, the flows are designed for best
effort services. There is no guarantee of data delivery or QoS requirement with
unspecified delivery bit rate and time.
● Service planning: Different from capacity planning, service planning targets at
services with specified QoS, such as data rate, latency, reliability, etc.
Individual flow
Backbone flow
Core
Composite flow
Composite flow
f low
te
osi
o mp
C
Individual flow
Composite flow
Figure 6.1 Flow types
There are basically following three flow types (cf. Figure 6.1):
● Individual flow: is the flow for a single session of an application. The individual
flow is devised for specified requirements.
● Composite flow: is the combination of individual flows that share the same path
and links within the same network. The composite flow is normally used in
capacity planning for best effort services.
● Backbone flow: is formed by composite flows to link different networks following
network hierarchy.
These above flow types have implications on capacity planning in the network. They
are useful for developing routing protocols and addressing schemes, especially in a
large network with a number of flows.
Identifying data flow types is important to help us select appropriate technology
and protocols. Such flow analysis process not only provides an end-to-end perspective
on link requirements but also establishes flow boundaries in the network hierarchy.
The flow analysis indeed allows us to gain some insights into the network hierarchy
degree and redundancy (if necessary). Useful information for determining intercon-
nection strategy should also be attained through the flow analysis. Switching, routing,
or hybrid techniques are expected to be developed with respect to different flow
specifications for either capacity or service plan.
Network flow analysis 117
6.2 Data sources and sinks

Discussing about flows, they always need to have two ends. One end is called source
(or transmitter or sender), while another end is called sink (or receiver or destination).
In the context of data communications in networks, the sources and sinks imply data
sources and data sinks, respectively, which can be defined as follows:
● Data source: is a node that produces data to convey to another node or multiple
nodes within the network. The data source can be either a network device or
a group of devices coordinated with each other in the data transmission. Some
examples of data sources are computing servers, parallel systems, and computing
clusters.
● Data sink: is a node that receives data from another node or multiple nodes. These
nodes play the role as the sources of this sink node. The data sink simply acts as
a data repository. Some examples of data sinks are display devices (e.g., monitor
and projector) and disks/tapes for data storage.
In practical networks, most devices can do both tasks of sending and receiving data.
Therefore, they act as either sources or sinks. Some can even do both the data trans-
mission and reception simultaneously, which is known as full-duplex system, while
others can support only half-duplex systems where the nodes can transmit and receive
data from each other but not simultaneously. A typical example of half-duplex device
is a walkie-talkie with a push-to-talk button allowing a person to either talk or listen
to another person, but not both at the same time. For full-duplex communications, the
telephone that we are using almost everyday can be taken as an example. The tele-
phone device with the employment of a pair of simplex links for two communication
paths allows people at both ends speak and hear at the same time.
As an initial step of the flow analysis, the components of the flows, i.e., the data
sources and sinks, need to be indicated to find out where flows are generated and
where they are going to. Then, we can create flow models for the network design and
analysis.
6.3 Flow models

Flows are generally variant in different network infrastructures depending on both
logical and physical designs. In order to simplify the flow analysis, the flows should
be first mapped into well-known flow models; for instance, peer-to-peer (P2P) model,
client–server model, cooperative computing model, and distributed computing model.
Flow models are characterized by the following:
● Directionality: The flows in practice are generally asymmetric. Each flow may
have its own requirements in each direction.
● Hierarchy: The degree of flows can be described through network hierarchy.
Users, hosts, addresses, and networks (subnetworks) are commonly grouped via
network hierarchy. Such hierarchical grouping helps in assigning appropriate
Figure 6.2 Peer-to-peer model
flows for each of these groups. Furthermore, it facilitates the identification of

critical flows which are the most important ones in the network, such as the
backbone flows to link different networks.
These two characteristics of the flows can be realized in different models which will
be discussed in the following subsections.
6.3.1 Peer-to-peer model

P2P model is applied when users or hosts and applications within the network are
likely similar. All users have the same access requirements of an application.
In the P2P model, all nodes play the same role, and thus, there is no obvious
directionality and hierarchy (cf. Figure 6.2). A typical example of the P2P flow
can be seen in tele-service applications (e.g., teleconferencing) where people with
different devices connect remotely with each other to share audio, video, and data via
a telecommunications system. The communication media could be over Internet or
LAN/WAN, and the supported devices could be computers, normal handsets, and/or
smart phones. The P2P model can also be deployed over the wireless in the form
of device-to-device (D2D) communication systems which have been developed as an
underlay to cellular networks. Various applications of the P2P as well as D2D systems
can be referred to the relevant works provided in Section 6.8.
Since there is no guarantee of data delivery over the P2P model, this model is
normally used in the capacity planning for best effort services.
Client
Client Server
Client
Client Client
Client
Figure 6.3 Client–server model
6.3.2 Client–server model

Client–server model is the most applicable flow model which has both flow char-
acteristics, i.e., directionality and hierarchy. In this model, the clients are people
who request the services, while the servers are service providers which could be web
server, application server, database server, mail server, etc. (cf. Figure 6.3). Examples
of client–server model can be found in the World Wide Web (WWW) on the Internet
and network printing within the LAN.
The communications in the client–server model are carried out via two ways
between the clients and the servers following the network hierarchy. One way is for the
request from the client, while the other is for the response from the server. Generally,
there are more responses consuming higher network resources than the requests,
which mean that the two-way data flow is asymmetric, mainly in the direction to the
clients. In a very large network where the response flows are much larger than the
request ones, such asymmetric data flow can be approximated as unidirectional flow.
6.3.3 Cooperative computing model

Cooperative computing model is an hierarchical flow model where applications col-
laboratively work and share data to accomplished a required task or service. There is
a manager with higher level application to control multiple client–server applications
(cf. Figure 6.4). The cooperative computing model is thus more hierarchical than the
client–server flow (cf. Section 6.3.2). Examples of cooperative computing model can
Manager
Server Server
Server-to-server
Client Client Client

Client Client Client
Client
Client
Figure 6.4 Cooperative computing model
be found in monitoring systems (e.g., weather/climate, water, etc.) via either wired
or wireless communications between sensors or computing nodes.
In the cooperative computing flow model, apart from server-to-client flows as in
client–server model, there are server-to-server flows to connect the servers for col-
laborative work and server-to-manager flows to connect the servers with the manager
at the top level.
6.3.4 Distributed computing model

Distributed computing model is a specialized flow model where the flows are mainly
either between computing nodes or between a manager and its computing nodes
(cf. Figure 6.5).
In distributed computing flow model, depending on the amount of information
transferred between computing nodes, we can categorize them as follows:
● Loosely coupled nodes: when these nodes have little or no information to be
transferred between them.
● Closely coupled nodes: when the information is frequently transferred across
these nodes.
The computing nodes are required to accomplish various computing tasks which
are classified as follows:
● Coarse granularity: Multiple tasks are assigned to the computing nodes in the
manner that each of them is dedicated to a single computing node.
● Fine granularity: A task is divided into several subtasks to be concurrently
manipulated at different computing nodes.
Task
manager
Client
Client
Computing nodes
Client
Client
Figure 6.5 Distributed computing model
Considering the categories of both computing nodes and tasks, following are the
two specified models related to distributed computing:
● Computing cluster: is a group of loosely coupled computing nodes to accomplish
a coarse-granularity task. In this case, the flows are asymmetric, which are mainly
from the nodes to task manager. Hence, the node-to-manager or node-to-server
flows are critical flows. Since the task for each node is independent of that for
another node, the flows in the computing cluster are not synchronized.
● Parallel computing system: is a set of closely coupled computing nodes to carry
out a fine-granularity task. Specifically, each task is divided into several subtasks
for computing nodes to concurrently work on along with continuously sharing
information between them. The communications between nodes is critical in the
parallel computing system; hence, node-to-node flows are critical flows.
In the above two computing models, flow characteristics vary depending on
the degree of granularity and coupling. Therefore, in order to optimize flow perfor-
mance, multicasting should be taken into account subject to the constraints on the
task requirements and capabilities of computing nodes.
6.4 Flow boundaries

In a large system consisting of a number of components connecting over multiple
networks (subnetworks), it is vital to define boundaries between different group of
components for simplifying network design and analysis.
LAN
LAN
MAN
LAN
LAN
WAN
LAN MAN
LAN
LAN
LAN
Figure 6.6 Flow boundaries
The flow boundaries are defined as the separations between large portions within
a system. They are helpful in indicating not only network flow hierarchies but also
directionality and consolidation of the flows. Flow boundaries are thus required to be
determined as part of the flow analysis.
A simple way to set up the boundaries is based on geographic areas, such as
between campuses, buildings, and floors. The network topology is also a factor
to define the flow boundaries, such as LAN/WAN, LAN/MAN, and MAN/MAN
separations (cf. Figure 6.6).
Another separation approach is to use logical boundaries to separate between
backbones, between network access points in the Internet, between WANs, or between
areas with different service requirements. For simplicity, service providers and net-
work engineers normally consider the network boundaries between the WANs in the
logical model.
6.5 Flow distributions

After determining flow boundaries between different networks, the next step in the
flow analysis is to assign flow distributions, through which the location of backbone
flows can be determined. As defined in Section 6.1, the backbone flows link different
networks over network hierarchy. A network engineer should first identify the regions
80% 80%
20%
LAN
LAN
WAN
20%
LAN
LAN
80%
80%
Figure 6.7 Flow distribution
for the flow distributions, e.g., LANs, MANs, and WANs, or more specifically with
campus area networks and home area networks.
The flow distributions with identified backbones are helpful in showing when
the flows stay in one region or transit from one region to another region. A typical
example of flow distribution is within a LAN and between LANs in a WAN. A simple
80/20 rule can be applied where 80% of the traffic flow is within the LAN and 20% is
across the WAN (cf. Figure 6.7). This means that the WAN capacity is approximately
25% of the LAN capacity. For example, 2.048 Mbps T1/E1 WAN circuit connects 10
Mbps LANs.
6.6 Flow specification

Along with the identified flows, boundaries, and their distributions, we need to build
a flow specification, a.k.a. flowspec, taking into account the requirements of each
application. Based on this collective requirement list, we can determine how to group
these requirements and assign them to the corresponding flows. The flowspec can be
then constructed.
In general, the flowspec has three forms as follows:
● One-part flowspec: is used for capacity planning to provide best effort services.
The flows are thus named best effort flows. There are no specified flows and no
guarantee of data delivery.
● Two-part flowspec: contains not only best effort flows but also specified or
predictable flows. This flowspec can be applied for both capacity and service
planning.
● Multipart flowspec: consists of all the above in addition to provide more details
and requirements on individual components of the flows.
Given the list of all flows and their characteristics, a flowspec algorithm can be
developed to bring all these characteristics together into a specification in the form of
combined flows. All characteristics for each flow, including capacity/throughput,
delay/latency, and RMA (if any available), should be combined to describe the
expected overall performance.
The development of flowspec algorithm should meet the following conditions:
● For best effort flows: only capacity requirements are involved in the calculation
for these flows.
● For specified flows: all performance characteristics (if available), i.e., capac-
ity/throughput, delay/latency, and RMA, are computed for these flows. For each
of these characteristics, it is aimed at solving an optimization problem to maximize
the overall performance of the specified flows.
● When delay and/or reliability requirements are indicated, these can be considered
as constraints in the optimization problem and should be treated individually in
the calculation for the flowspec.
● Capacities produced by one-part, two-part, and multipart flowspecs are regarded
as baseline performance only, while the performance modifiers rely on various
factors in practical network deployment.
The flow characteristics in flowspec algorithm can be computed for different
flowspec types as follows:
● One-part flowspec: Considering capacity planning with L best effort (BE) flows,
let CBE,i , i = 1, 2, . . . , L, denote the capacity of the ith flow. The total capacity
CBE is given by

L
CBE = CBE,i . (6.1)
i=1
● Two-part flowspec: Considering both capacity and service planning with L

BE flows and K specified (SP) flows, let CBE,i , i = 1, 2, . . . , L, and CSP, j ,
j = 1, 2, . . . , K, denote the capacity of the ith BE flow and the jth SP flow, respec-
tively. The total capacity for BE services, i.e., CBE , is given as in (6.2), while the
total capacity in specified environment, i.e., CSP , is computed by

K
CSP = CSP,j . (6.2)
j=1
In addition, the reliability and delay of the SP flows denoted by RSP and DSP ,
respectively, should be considered in two-part flowspec.
● Multipart flowspec: Along with performance characteristics devised as in one-

part and two-part flowspecs, the specified services of the jth flow, j = 1, 2, . . . , K,
should be guaranteed at the minimum of RSP,j , DSP,j , and CSP,j .
6.7 Flow analysis model

The above flow analyses with identified models, distributions, characteristics, and
flowspec algorithm are all combined into developing guidelines for the flow analysis
which basically consists of the following steps (cf. Figure 6.8):
● Step 1—Applying flow models: In this initial step, application map is first built.
Then, flow model is developed, in which data sources and sinks are identified
along with their locations, flow directionality, and hierarchy.
● Step 2—Establishing flow boundaries: Based on the identified flow model, data
sources, and sinks, this step sets up flow boundaries and identifies which flows
cross these boundaries. In particular, critical flows are determined among those
crossing the boundaries.
Identifying flow models
Establishing flow
boundaries Boundary?
Determining flow
distributions Traffic?
Combining flow models,

boundaries, and
distributions
Data aggregation and analysis
Developing flowspec
Prioritizing flows
Figure 6.8 Flow analysis model

● Step 3—Applying flow distributions: Determining flow distributions is critical

in network design and optimization. The flow distributions show the expected
traffic volumes of a flow when it crosses the boundaries. This step is also helpful
in determining the necessity of generating a backbone flow across the network.
There are some specific guidelines with respect to flow models as follows:
– P2P model: Due to not having obvious flow directionality and hierarchy, there
is no particular flow distribution in this model.
– Client–server model: Flow distribution 50/50 can be applied for the case
when clients cross a boundary to reach a server, otherwise flow distribution
80/20 is used as in the LAN/WAN separation.
– Cooperative computing model: Flow distribution 50/50 can be applied for the
case when the flows from different servers must cross a boundary to arrive at
a server, otherwise applying 80/20 rule.
– Distributed computing model with computing cluster: Flow distribution 50/50
is applied since the computing nodes are remote from the task manager, and
all of them have to cross a boundary.
– Distributed computing model with parallel computing system: Flow distribu-
tion 20/80 is applied since all the traffic flows between computing nodes are
across flow boundaries.
Note that the above guidelines can be changed to adapt with practical network
design given different requirements of applications, host and user distribution,
network topology/model, etc.
● Step 4—Combining flow models, boundaries, and distributions: This step sim-
ply combines the information collected in the previous three steps to indicate
the likelihood of making use of backbone flows and their location in network
hierarchy.
● Step 5—Developing flowspec: Summarizing all the above steps, this step helps
in providing capacity and service plans along with requesting, configuring, and
verifying services. These also facilitate the selection of appropriate technology to
meet service requirements. In order to develop the flowspec, as outlined in Section
6.6, we need to collect a list of applications and their requirements, application
map, data sources and sinks, flow models, boundaries, distributions.
● Step 6—Prioritizing flows: The flowspec developed in Step 5 enables us to
prioritize flows according to the number of users, distance, location, and the
performance characteristics of each flow.
6.8 Related works

references:
Information Flows Ford in 1956 [29], Ford and Fulkerson in 1956 [30], Omidyar
and Pujolle in 1996 [122], Brownlee et al. in 1999 [123], and Floyd and Fall in
1999 [124], Gamal and Kim in 2012 [125], and Frana in 2018 [126].
P2P Model Shen et al. in 2009 [127], Fitzek and Charaf in 2009 [128], Gurler and
Tekalp in 2013 [129], Kumar and Lee in 2014 [130], Yin et al. in 2017 [131],
Wu et al. in 2017 [132], and Jiang et al. in 2018 [133].
D2D Systems Asadi et al. in 2014 [134], Trestian et al. in 2015 [135], Wu et al. in
2016 [136], and Ali et al. in 2016, 2017, and 2018 [137–139].
Client–Server Model Sinha in 1992 [140], Ramesh and Perros in 2000 [141],
Kanitkar and Delis in 2002 [142], AbdulFatah and Majumdar in 2002 [143],
and Zhu et al. in 2016 [144].
Cooperative Computing Taylor in 1990 [145], Richardson et al. in 1998 [146],
Wu et al. in 2013 [147], You and Huang in 2018 [148], and Sheng et al. in
2018 [149].
Distributed Computing Casavant and Kuhl in 1988 [150], Morse in 1994 [151],
Bianchi in 2000 [152], Younis and Fahmy in 2004 [153], Nishida and Nguyen in
2013 [154], Kandaswamy et al. in 2018 [155], and Pascale et al. in 2018 [156].
Flow Boundaries Estrin et al. in 1989 [157] and Quoitin et al. in 2003 [158].
Flow Distributions McCabe in 2007 [73] and Oppenheimer in 2010 [77].
Flow Specification Partridge in 1992 [159] and Marques et al. in 2009 [160].
Flow Analysis and Control Cruz in 1991 [161], Parekh and Gallager in 1993 [162],
Jane et al. in 1993 [163], Sen and Wang in 2004 [164], Popovski and Yomo in
2007 [165], Katti et al. in 2008 [166], Sohn and Kim in 2008 [167], and Liu and
Chen in 2015 [168].
Flow Virtualization Dally in 1992 [169], Jagadeesan et al. in 2014 [170], Lara et al.
in 2014 [171], Kreutz et al. in 2015 [172], Hassan et al. in 2017 [173], Li et al.
in 2018 [174], and Jia et al. in 2018 [175].

[Q6.1] What is a flow between two end nodes?
[Q6.2] Describe three common flow types within a network.
[Q6.3] Give an example of data sources and data sinks for each of the following
kinds of end nodes:
● The end nodes are data sources only.
● The end nodes are data sinks only.
● The end nodes are either data sources or data sinks.
[Q6.4] How to characterize flow models?
[Q6.5] What is peer-to-peer (P2P) model?
[Q6.6] Describe the communications in client–server model.
[Q6.7] What kinds of flows in cooperative computing model?
[Q6.8] What are loosely coupled nodes and closely coupled nodes in distributed
computing model?
[Q6.9] Describe two kinds of computing tasks at nodes in distributed computing
model.
[Q6.10] Differentiate between cooperative computing and distributed computing
models.
[Q6.11] Explain the usage of three flowspec forms.

[Q6.12] What conditions are required in the development of flowspec algorithm?
[Q6.13] Explain steps in a flow analysis model and show how to determine the
appropriate flow distribution in different networks.
6.10 Problems
[P6.1] Discuss how to design flows for capacity planning and service planning
schemes in an enterprise subject to various service requirements in different
areas/buildings/offices.
[P6.2] Provide an example of a communication network where peer-to-peer model
is incorporated in client–server model.
[P6.3] Assume the servers are located in the cloud. This computing model is also
known as cloud computing. Do the research on this model and show how
to adapt cooperative computing and distributed computing models for the
cloud-based servers.
[P6.4] Apart from the popular 80/20 rule, investigate other flow distribution rules,
e.g., 50/50, 20/80, etc., and their applications in practical networks.
[P6.5] In the flow analysis model, discuss criteria for prioritizing flows based
on flowspec. Show their corresponding impacts on the network flow
performance in the whole network.
Chapter 7
Network performance evaluation
In the previous chapters, we have gone through the whole network-design process
starting from the initial requirement analysis to flow analysis with detailed guide-
lines for configuring network elements and entire network, as well as mechanisms
for developing network and flow specification. In order to validate the effective-
ness of a network design, its performance needs to be evaluated. In general, there
are three approaches to evaluate the network performance, including benchmarking,
simulation, and analytical modeling.
In this chapter, I will sequentially discuss these three approaches along with their
merits and demerits as well as their applications. Section 7.1 will first present the
benchmarking technique, followed by the introduction of simulation and analytical
modeling approaches in Sections 7.2 and 7.3, respectively. Also, in this chapter, a
typical system and its component will be provided in Section 7.4 as illustrations of
different performance-evaluation techniques.
7.1 Benchmarking: merits and demerits

Benchmarking is an evaluation technique which has been well employed in various
network models to measure network performance while a standard set of task or
workload is running. Taking into account different approaches and solutions over
a number of tests and trials, the benchmarking process in particular allows a fair
comparison between them.
The benchmarking technique is helpful in building a new system and also when
modifying the existing system. Some of its merits can be intuitively observed as
follows:
● Reliability: With a number of trial runs given various standard sets, the results
obtained with benchmarking are shown to be more accurate than other approaches.
● Applicability: After a rigorous validation, the benchmarking results can be directly
applied in practice.
● Understandability: As a practical approach, the benchmarking is easy to grasp.
● Marketability: Due to its practicability, the benchmarking approach can be easily
marketed to be sold.
However, apart from these above merits, the benchmarking technique has some
essential demerits as follows:
● Cost: The benchmarking requires a high cost in building the system, instrumen-
tation, and testing before the product is able to come into use.
● Time: Such high complicated process with benchmarking technique is definitely
time-consuming.
● Personnel: The success of the product after a number of tests and trial runs requires
a lot of efforts, and human factor plays a key role in the benchmarking process.
● Extendability: Some changes in the system or network environment may have
considerable impacts on the results. The benchmarking results are hard to be
extended for different cases; hence, different benchmarking processes are needed.
7.2 Simulation: merits and demerits

Simulation is a well-known approach for modeling a system as well as evaluating
network performance. We can build a simulation model of the system under investi-
gation by using various existing simulators, such as Riverbed Modeler, discrete-event
network simulator with NS3 for Internet systems, or OMNeT++ based on C++
programming language.
The simulation approach can provide a simple modeling of the network for assess-
ing its performance and generally does not require high mathematical skills. The
simulation has shown to provide a number of advantages as follows:
● Adaptability: The simulation program can be run and rerun with several alternative
parameter sets to obtain various results for validation.
● Flexibility: The simulation is flexible in the sense that the simulation models
allow us to collect results at different stages during the simulation process to keep
track of all generated data.
● Extendability: With the above merits of flexibility and adaptability, the simulation
models are generally easy to extend for different networks.
Although the simulation has shown a number of merits, it has some demerits that
need to be considered in the performance-evaluation process, such as the following:
● Personnel: In order to construct a new simulator or simulation software, it nor-
mally requires a lot of human efforts, time, and work commitment, especially
with inexperienced engineer or a programmer with no specialized knowledge in
that simulator.
● Time: The simulation is shown to be much faster than benchmarking. However,
a large number of simulation runs are required to elaborate and validate the
correctness of the results, which is a time-consuming process depending on the
length of the simulation program. Although the simulation programs are easy to
run and rerun for a number of times, very minor changes of the system parameters
may require new simulation runs.
Network performance evaluation 131
● Managing: Some simulation programs are prohibitively bulky with a number of

functions, controls, variables, etc., Understanding all of these may be easy, but
managing them is a challenging task that every programmer is facing. A manage-
able approach should be considered at the beginning of programming or writing
with any simulator.
7.3 Analytical modeling: merits and demerits

Analytical modeling, a.k.a. mathematical modeling, has been well established to pro-
vide an abstract model of a network or system instead of performing simulation with
various simulators or experiment with hardware implementation. The mathematical
model allows us to provide exact or approximate solution of a problem, such as eval-
uating channel capacity, network or system throughput, latency, reliability, security,
etc. Some of these, e.g., reliability, latency, and capacity, have been discussed in
Section 5.2.
In order to analyze network performance via mathematical modeling, cutting-
edge mathematical techniques and tools can be employed, which are shown to be
very computationally efficient. Different mathematical methods can be applied, such
as differential/integral calculus, algebraic methods, probability theory, stochastic
process, queuing theory, etc. In particular, these techniques can be easily employed
for several cases when the model has been validated. Due to its convenience, the
analytical modeling has emerged as a promising approach.
Specifically, the mathematical modeling has some merits as follows:
● Computability: A good mathematical modeling should provide a fast computation

of various performance measures. Different mathematical tools or software, e.g.,
MATLAB or MATHEMATICA, can also be used for the computation.
● Adaptability: The derived expressions through the mathematical analysis can be
easily employed to achieve the results for different cases.
● Optimality: Based on the derived expression of the performance measures, we can
maximize the network performance. Various optimization problems can be devel-
oped to find the optimal network parameters subject to constraints in accordance
with requirements which could be from different perspectives.
The analytical modeling, however, has some disadvantages due to its inherent
restrictions and requirements as follows:
● Mathematical skills: High-level mathematical skills are required to derive the

expressions of various performance metrics which could be in either approximated
or closed forms.
● Artificiality: Several assumptions are considered in analytical modeling to sim-
plify the analysis. These assumptions are sometimes too artificial and unrealistic
to be able to apply for practical cases.
Attribute(s) State
Entity
Exogenous events
Activity
Endogenous events
System
Figure 7.1 System components in a simulation model
● Generalization: The analytical modeling can generalize the problem with a

number of formulas. However, it does not provide detailed information with
specification as in simulation approach.
● Accuracy: Due to various assumptions, the results obtained by the analytical
modeling are not as accurate as in benchmarking.
● Practicality: Exploiting the mathematical formulas, the derived expressions can
provide a quick validation of the performance, but some of them can only be
derived in approximated forms. With the evolved complex analytical models, it is
promising that these problems can be solved in a general closed form to be able
to compare with those in practice.
7.4 Systems and system environment

The above performance-evaluation techniques can be applied to either each compo-
nent of the system or the entire system. In order to understand the applications of
these techniques, this section will introduce some basic concepts of systems and their
components.
In the context of network performance evaluation, a system is defined as a group
of objects that are working together in a collaborative manner to accomplish several
tasks. Any changes in system environment may affect the behavior of the system. It
is therefore vital to understand the functionality and operation of the system and its
environment.
In general, a system has the following components (cf. Figure 7.1):
● Entity: is an object within a system.
● Attribute: is a property of an entity.
Data packets
Data flow
Transmitter Receiver
Figure 7.2 Data flow between two end devices
● Activity: is an action over a specified period of time.

● State: is a collection of related variables which is necessary to describe the system.
The state in general is a function of time. In some systems, for simplicity in the
analysis, the state is assumed to be quasi-static, which is invariant for a fixed
period of time.
● Event: is an instantaneous occurrence of an entity. The event may result in a
change of the state.
The activities and events causing the changes of the system could have either
internal or external causes. These two kinds of activities and events are defined with
the following terms:
● Exogenous activities and events: refer to the activities and events in the
environment that affect the system.
● Endogenous activities and events: refer to the activities and events occurring
within a system.
Considering the change of state over time, a system can be classified into the
following two types:
● Discrete system: is a system in which the state variables change in a discrete set
of points over time.
● Continuous system: is a system in which the state variables change continuously
over time.
Although these two kinds of systems are clearly defined, they may be used inter-
changeably depending on the state variables under consideration. For example, in a
communication system, the number of packets arrived is discrete, but the data flow
is continuous (cf. Figure 7.2).
Example 7.1. A typical example of the above concepts can be found in a bank (cf.
Figure 7.3). In the bank, every customer is an entity, and account balance is an
attribute of the considered entity. The customer is assumed to go to the bank to make
deposits. In this case, making deposits is an activity of the entity. The arrivals of
the customer and the completion of the services are events in the system. To describe
the processing of the system, the number of the customers waiting in the queue and the
State:
Attribute: The number of
Account balance customers waiting
and waiting time
Customer
Event: Arrival of customer

Activity:
Deposit money
Event: Completion of service
Banking system
Figure 7.3 Components in a banking system
State:
The number of
Attribute: packets waiting in
Packet length/size the buffer and
waiting time
Data packet
Activity: Event: Arrival of packet

Transmission and
reception of data Event: Processing of packet
Single-server
queuing system
Figure 7.4 Components in a single-server queuing system
waiting time are state variables. The arrival of the customer is an exogenous event,
while the completion of the service for the customer is an endogenous event.
Example 7.2. Analogically, consider a single-server queuing system for data com-
munications (cf. Figure 7.4). The entities in this system are the packets, and their
length is an attribute. The transmission and reception of the packets are activities of
these entities. In this case, the event can be the arrival of the packets at the server,
and the state variable is the number of packets waiting in the queue to be sent. Details
of the queuing system will be discussed in Chapter 15.
7.5 Related works

references:
Benchmarking Gray in 1992 [176], Madhavan et al. in 2009 [177], Harcho-Balter
in 2013 [178], and Nambiar and Poess in 2016 [179].
Analytical Modeling Wolfram Mathematica [180], MATLAB [181], Abell and
Braselton in 2017 [182], and Wolfram in 2017 [183].
Simulation and System Components Varga in 1999 [184], Banks et al. in 2010
[185], Wehrle et al. in 2010 [186], Davis in 2010 [187], Rimoldi in 2016 [188],
Choi and Liu in 2016 [189], Attaway in 2017 [190], NS3 [191], OMNeT++ [192],
MATLAB [181], and Simulink [193].
[Q7.1] What is benchmarking? What are its advantages and disadvantages?

[Q7.2] What are the advantages and disadvantages of simulation in general?
[Q7.3] How to perform analytical modeling and what skills are indispensable?
[Q7.4] Discuss the merits and demerits of the analytical modeling.
[Q7.5] List components in a simulation model and explain each of them in the
context of a computer network.
[Q7.6] What are exogenous and endogenous events and activities?
[Q7.7] Differentiate between discrete and continuous systems. Provide an example
for each kind of these systems.
[Q7.8] Describe the components within a banking system and a single-server
queuing system. Illustrate their relationship in a diagram for each system.
7.7 Problems
[P7.1] Select a network-simulation software. Discuss about its advantages and
disadvantages in modeling a real system.
[P7.2] Compare the complexity of analytical modeling and simulation approaches.
Any simulation software or tool can be selected.
[P7.3] Discuss the accuracy of the results derived from analytical modeling
compared with those obtained from simulation software.
[P7.4] Provide an example to show the artificiality and impracticality of the
analytical modeling.
[P7.5] Build a table comparing merits and demerits of three approaches to evaluate
network performance.
[P7.6] Develop a simulation model for a multi-server queuing system. Illustrate its
components and their relationship in a diagram.
Chapter 8
Network simulation models
In the previous chapter, we have learnt different techniques for network-performance

evaluation. Every technique has its own merits and demerits. Depending on specific
requirements and capabilities, we can select one or a combination of multiple tech-
niques. Specifically, simulation is the most popular approach due to its simplicity,
flexibility, and adaptability with a number of tools and software. This chapter will
discuss in detail simulation models for emulating and evaluating a system.
In this chapter, I will first introduce simulation techniques used for emulating
and evaluating a system in Section 8.1. The selection of appropriate simulation tools
along with their advantages and disadvantages will be also discussed. Simulation
models of a typical system will be presented in Section 8.2. Detailed steps and phases
of a simulation study will be outlined in Section 8.3. Simulation examples will be
then provided in Section 8.4 to show the effectiveness of network simulation models.
8.1 Simulation for emulating and evaluating a system

Simulation is well known as the imitation of the operation of a system in real world. A
simulation model is developed to study the behavior and characteristics of the system.
Prior to being deployed, the simulation model needs to be validated. After proving
validity and accuracy, the simulation model can be employed as an analysis tool. It
can be used to study the system at the design stage as if in a practical system, allowing
us to evaluate various performance metrics of the system in different circumstances.
Some applications of simulation can be found in manufacturing systems, construc-
tion engineering, military, transportation, traffic, health care, logistics, business and
project management, etc.
As discussed in Chapter 7, there exist different approaches for evaluating network
performance. The question is when simulation should be an appropriate tool. In order
to answer this question, it is crucial to know the benefits as well as potential application
of simulation for different purposes.
Specifically, a simulation model should
● enable experimentation with a system or a subsystem,
● simulate changes and evaluate the effects of these changes on system behavior,
● allow us to gain knowledge to improve the system,
● give a better understanding of system variables and their interaction,

● be used to verify the findings with analytical approaches or methodologies,
● be employed and tested prior to implementing a new system,
● allow us to investigate different capabilities of a system,
● be developed for different learning purposes,
● support visualization control and/or produce animation shows, and
● be exploited to treat some interactions within a complex system where a real
implementation would be challenging.
In fact, the above scenarios when employing simulation would be benefited due to a
number of its advantages. The simulation could
● allow new policies, procedures, algorithms, etc. that can be tested and validated
without interrupting ongoing process and operation of a real system;
● help us to test new hardware/systems to verify their functionalities and compati-
bility and/or to validate their performance prior to buying them;
● test various hypotheses on certain phenomena that can occur during the operation;
● allow us to either speed up or slow down the phenomena by compressing or
expanding the simulation time;
● provide insight about the interaction of variables within the system along with
their impacts on the system performance;
● perform bottleneck analyses of the system; and
● help us to understand the operation of an entire system by answering various
“What If ” questions.
However, the simulation is sometimes not appropriate in certain circumstances

when
● the solution of the problem can be intuitively observed by common sense,

● the problem can be easily solved via analytical method,
● it is easier to perform benchmarking technique,
● simulation software/tools are too expensive to invest in,
● resources and time are limited or even not available due to human and/or system,
● data/parameters required for the simulation are not able to be obtained,
● system behavior is too complex to perform simulation, and
● expected output of the simulation is unreasonable and infeasible.
The above limitation is due to the following disadvantages of the simulation, including
the following:
● Special training and experiences may be required for specific software.

● Some simulation results can be difficult to interpret for their meanings and
reasoning.
● Learning new simulation software is time-consuming.
● Simulation software is too expensive, while analytical method is not too hard to
implement.
Network simulation models 139
8.2 Simulation models of a system

As discussed in Section 7.4, a system is a group of objects working together to
accomplish several tasks. In simulation, the system needs to have a model which
is simply defined as a representation of the system. Such modeling is helpful for
studying the system.
A model is designed for one or several goals in the simulation. Therefore, we
may have different simulation models of the same system for different objectives; for
instance, when evaluating different performance metrics over different timescale of
the same system. While the validation of a real system could be very complicated in
practice, its simulation model would facilitate this process by simplifying the system.
However, the simulation model needs to maintain sufficient details to represent the
real system allowing us to draw valid performance evaluation results.
The simulation model is regarded as a special type of mathematical models which
use symbolic notation and mathematical equations to represent a system. Indeed, the
simulation model is usually exploited to validate various performance metrics of the
system which are expressed in the form of mathematical equations.
In order to verify and validate a simulation model, a number of tests or simulation
runs are required. Considering the change of the model over time, there are two types
of simulation models, including the following:
● Static simulation models: These models are used to represent a system at a

particular point of time.
● Dynamic simulation models: These models are used to represent systems as they
change over time.
For simplifying the modeling and analysis, the system can be assumed to be invariant
within a specific period of time. In this case, it is regarded as a static simulation
model.
Within a system, there may have one or several variables affecting the perfor-
mance of the system. The simulation models should represent these variables and thus
can be classified into two other kinds as follows:
● Deterministic simulation models: In these models, there is no random variable,

and all system parameters are invariant over time.
● Stochastic simulation models: These models contain one or multiple random
variables as inputs. In the simulation, these variables are randomly generated in
every run. Such random changes can imitate the behavior of a real system over
time, allowing us to provide statistical estimate of the true characteristics of the
system.
Considering the change of state over time of the random variables within a system,
there are also two kinds of the simulation models, including the following:
● Discrete simulation models: are used to represent the discrete systems in which
the state variables change in a discrete set of points over time.
● Continuous simulation models: are used to represent the systems in which the
state variables change continuously over time.
A typical model that is frequently considered in network design is “discrete,
dynamic, and stochastic simulation model.” In this model, there is one or multiple
random variables which change in a discrete set of points over time.
8.3 Steps/phases in a simulation

A simulation study basically consists of the following steps:
● Step 1: Formulate the problem that needs to be solved.
● Step 2: Set objectives and overall project plan.
● Step 3: Form model concept to meet the proposed plan and objectives.
● Step 4: Collect data required for simulation model.
● Step 5: Translate the model with simulation software/toolbox.
● Step 6: Verify the simulation model. If the model is verified, then go to Step 7.
Otherwise, go back to Step 5 to check the model-translation process.
● Step 7: Validate the simulation model. If the model is shown to be able to provide
the desired results, then go to Step 8. Otherwise, there may be something wrong
in the model conceptualization in Step 3 or the data collection in Step 4.
● Step 8: Design experiment to evaluate various performance metrics.
● Step 9: Run experiment and analyze the system performance.
● Step 10: Verify if the number of simulation runs is enough to represent the real
statistics of the considered performance metrics. If the number of runs is enough,
then go to Step 11. Otherwise, go back to Step 9 to run more simulations. In
case that more changes in the design are required, go back to Step 8 to amend the
experimental design.
● Step 11: Write report and analyze the results for the entire system given the
collected experimental data.
● Step 12: Implement the design.
For simplicity, the above 12 steps can be grouped into four phases (cf. Figure 8.1)
as follows:
● Phase 1—Formulation and planning: consists of Steps 1 and 2.
● Phase 2—Model construction: consists of Steps 3–7.
● Phase 3—Model running and data collection: consists of Steps 8–10.
● Phase 4—Data analysis, documentation, and implementation: consists of Steps
11 and 12.
8.4 An example of simulation model

In order to understand further the steps in a simulation study, let us consider an
example of a discrete simulation model to represent a queuing system.
Formulation and planning
Model construction
Model running
and
data collection
Data analysis,
documentation, and
implementation
Figure 8.1 Simulation phases
A queuing system is described by calling population, nature of arrivals, service

mechanism, system capacity, and queuing discipline (details of queuing system will be
discussed in Chapter 15). Specifically, a single-channel queue is considered, having
an infinite calling population, random arrivals, random service, unlimited system
capacity, and first-come first-serve queuing discipline.
In this simulation, the queuing system consists of a server, data packets being
served and data packets waiting in the queue to be served. The state of the system is
the number of data packets in the system, and the status of the server is either busy
or idle. An event is a set of circumstances that causes an instantaneous change in the
state of the system, e.g., arrival and departure events (cf. Example 7.2 in Chapter 7).
There is a simulation clock to record the duration of the simulation as well as the
number of simulation runs.
The simulation is carried out as in Figure 8.2 with the following phases:
● Phase 1: Determine the characteristics of each of the inputs to the simulation
model. Then, formulate the problem with objectives and constraints (if any) for
the simulation.
● Phase 2: Construct a simulation model consisting of p inputs xi , i = 1, 2, . . . , p,
and the corresponding response yi .
● Phase 3: Run simulation. For each simulation run, generate a value for each of
the p inputs and evaluate the function to calculate the response y.
● Phase 4: Aggregate all simulation data for writing report prior to implementing
for a real system.
Formulate an
optimization problem:
objectives and constraints
Construct a simulation
model with inputs and
corresponding responses
Run simulation with

different parameter sets
and variable generations
Aggregate simulation
data, plot figures, and write
report
Figure 8.2 Simulation phases in a queuing system
8.5 Related works

references:
Simulation Models and Process Jain in 1991 [114], Walrand et al. in 1999 [194],
Robinson in 2004 [195], Banks et al. in 2010 [185], Wehrle et al. in 2010 [186],
and Murray-Smith in 2012 [196].
Deterministic Simulation Models Kumar and Davidson in 1978 [197], Poole and
Raftery in 2000 [198], and Kleijnen in 2007 [199].
Stochastic Simulation Models Tijms in 1986 [200], Ripley in 1987 [201], Kulkarni
in 1995 [202], and Pinsky and Karlin in 2011 [203].
Discrete Simulation Models Varga in 1999 [184], Fishman in 2001 [204], Banks
et al. in 2010 [185], and Choi and Kang in 2013 [205].
Continuous Simulation Models Zeigler et al. in 2000 [206], Cellier and Kofman in
2006 [207], Birta andArbez in 2007 [208], and Bandyopadhyay and Bhattacharya
in 2014 [209].

[Q8.1] Discuss when simulation should be selected for modeling a system and
evaluating its performance.
[Q8.2] What are limitations and disadvantages of simulation?
[Q8.3] Differentiate between static and dynamic simulation models.

[Q8.4] In what systems the simulation models are considered as stochastic?
[Q8.5] Based on what criteria a simulation model can be categorized as discrete or
continuous?
[Q8.6] List all steps and phases in a simulation process.
[Q8.7] Provide an example of a simulation model experiencing all of the
steps/phases.
8.7 Problems
[P8.1] Provide an example for each type of the following simulation models:
● Static simulation model
● Dynamic simulation model
● Deterministic simulation model
● Stochastic simulation model
● Discrete simulation model
● Continuous simulation model.
[P8.2] Produce a table summarizing all the above simulation models highlighting
the differences between them.
[P8.3] Find a research article on networking which uses simulation to verify
the mathematical analysis. Do the research on the considered simulation
approach to show its necessity and effectiveness in network-performance
evaluation.
[P8.4] Provide an example of a complex system that simulation should be performed
prior to implementing in a real system.
[P8.5] Discuss on how to perform bottleneck analyses of the system with simulation.
[P8.6] Investigate why the simulation outputs are sometimes regarded as unfeasible
and/or unreliable in practical networks.
[P8.7] In a simulation model, how to verify if the number of simulation runs is
enough to represent the real statistics of the performance metrics under
investigation, i.e., Step 10 in the simulation process? Provide an example of
a simulation model where the number of runs is not large enough to validate
its performance.
Chapter 9
Statistical models in network simulation
In practical networks, most of the parameters are random and variant over time.
Statistical models are crucial in network simulation to represent the real network.
Specifically, queuing systems have been developed as a statistical model of the waiting
lines in telecommunication networks.
In this chapter, I will first introduce statistical models used for modeling queu-
ing systems with different distribution types in Section 9.1. Section 9.2 will present
Poisson point process (PPP) and its variants including homogeneous PPP and inhomo-
geneous PPP which have been well developed to model the number of events occurred
in the statistical models. This chapter will be closed by outlining detailed steps for
developing input models in Section 9.3, which include data collection, distribution
identification, parameter estimation, and testing.
9.1 Statistical models in queuing systems

Queuing models have been constructed for a number of applications in different areas,
such as telecommunications, computing, industrial engineering, data management,
etc. In the context of telecommunication networks, the queuing models will hereafter
consider the queuing of data packets.
In a queuing system, the interarrival time of data packets and service time at
servers are generally random. In order to model the queuing system in a probabilistic
way, some statistical models have been developed for the interarrival and service time
distribution as follows:
● Exponential distribution: If the duration is completely random, it is generally
assumed to follow an exponential distribution which is used to represent the time
between events in a PPP.
● Normal distribution: If the duration is fairly constant but with some ran-
dom variability which can be either longer or shorter, it can follow a normal
distribution.
● Truncated normal distribution: If the duration also follows normal distribution
but it is restricted within a specific interval, it can be modeled by truncated normal
distribution with bounded random variable.
● Gamma andWeibull distribution: This distribution is a general form of exponential

distribution when both the shape and scale are taken into account.
Details of the above distributions and their properties will be discussed in the fol-
lowing chapters. In addition, the number of packets arrived in the system during a
specific interval follows a Poisson distribution, a.k.a. PPP, which will be discussed in
Section 9.2.
For evaluating the reliability and maintainability, the time to failure is another
random variable that needs to be considered. In order to represent this, some statistical
models can be used, such as follows:
● Exponential distribution: in case that the failures are random.

● Normal distribution: in case that the failures are due to wear after continuous use.
● Gamma and Erlang distribution: is generalized from exponential distribution to
characterize standby redundancy for failures where each component follows an
exponential distribution.
● Weibull distribution: to model failures caused by a large number of defects in a
system.
For applications in different areas, in order to model a queuing system, various

distributions can be employed. For example, in inventory and supply chain, Gamma
distribution is used to model the lead time. The number of units demanded per order
can follow either Poisson distribution for a general demand or negative binomial
distribution, a.k.a. Pascal distribution, for a larger number of demands or geomet-
ric distribution as a special case of the negative binomial distribution when at least
one demand should occur. There are some other useful distributions that can be found
in different applications, such as Bernoulli, binomial, Erlang, uniform, triangular,
beta, etc.
The above random phenomena can be either discrete or continuous over time,
and thus, the distributions can be classified as
● Discrete distributions: are used to describe random phenomena having only

countable values.
● Continuous distributions: are used to describe random phenomena in which the
variable can take an infinite and uncountable value.
Apart from the well-known distributions, there are empirical distributions in

which the parameters are observed values from experiment with data sampling. These
distributions are helpful when it is impossible to figure out the exact or approximated
statistical distribution of the random phenomena. Since this approach is based on
the practical experiment, its inherent advantage is that there is no assumption when
observing and analyzing the data. However, a number of samples are required for an
entire range of possibilities, which is sometimes unfeasible to cover all cases.
Statistical models in network simulation 147
9.2 Poisson point process

PPP has been well studied in probability and statistics field to model the number of
points randomly located in Euclidean space which can be in one, two, three, or higher
dimensions. In queuing theory, the PPP is defined in one dimension, i.e., on a line,
as a stochastic process to represent a counting process in a queuing model, while
in higher dimensions, it can be used in wireless networks to model the locations of
scattered objects, e.g., transmitters and receivers, in space.
The PPP is a special case of a more general process known as Markov process or
birth–death process. The Markov process is a stochastic or random process in which
the future probabilities are determined by the current state of the process but not on
how that state was reached. The Markov process is characterized as “memoryless.”
As will be shown later, the PPP also has the memoryless property of the Markov
process.
In order to understand the PPP, Poisson random variable and its distribution will be
discussed in this section. The Poisson distribution can be defined as in Definition 9.1.
Definition 9.1. The Poisson distribution is a discrete distribution of the number of
events that occur randomly in a given interval of time or space.
Let us denote the number of events occurring in a given interval and its average
value by X and λ, respectively. Here, λ is also known as the event rate or rate parameter
of Poisson distribution. We have the following definition of the Poisson random
variable.
Definition 9.2. X is a Poisson random variable if the probability of observing x,
x ∈ N, events in a specified interval is given by
e−λ λx
Pr [X = x] = , (9.1)
x!
where x! denotes the factorial of x computed by

x
x! k. (9.2)
k=1
Notice that 0! = 1, and thus, Pr [X = 0] still exists which is given by

Pr [X = 0] = e−λ .
The Poisson random variable has the same average value, a.k.a. mean or expected
value, and variance which are given by
E[X ] = V [X ] = λ, (9.3)
where E[·] denotes the expectation operator and V [·] denotes the variance operator.
Let N (t) denote a counting function that represents the number of events occurred
in [0, t]. If the event rate, i.e., λ, is a constant, we have homogeneous PPP or stationary
PPP (SPPP). Otherwise, we have inhomogeneous PPP or nonstationary PPP (NPPP).
In the following subsections, the SPPP, NPPP, and their operations will be sequentially
presented in Sections 9.2.1–9.2.3.
9.2.1 Stationary Poisson point process

Definition 9.3. A counting process {N (t), t ≥ 0} is a homogeneous PPP or SPPP
with mean rate λ if
● Arrivals occur one at a time.
● N (t) has stationary and independent increments.
● N (t) follows Poisson distribution having the following probability function
e−λt (λt)n
Pr [N (t) = n] = ,n ∈ N (9.4)
n!
Similar to Poisson random variable, N (t) has equal mean and variance which are
given by
E[N (t)] = V [N (t)] = λt. (9.5)
Due to the stationary increment, the number of events in the interval (a, b], a ≤ b,
denoted by N (a, b] follows Poisson distribution with
e−λ(b−a) (λ(b − a))n
Pr [N (a, b] = n] = , n ∈ N. (9.6)
n!
The average number of events in (a, b] is given by
E[N (a, b]] = λ(b − a). (9.7)
The number of events is also independent in any disjoint intervals, i.e., (a1 , b1 ], (a2 , b2 ],
. . . , (aK , bK ], and thus
Pr [{N (a1 , b1 ] = n1 , N (a2 , b2 ] = n2 , . . . , N (aK , bK ] = nK }]

K
e−λ(bi −ai ) (λ(bi − ai ))n
= i
. (9.8)
i=1
ni !
Example 9.1. Packet arrivals at the server occur randomly following a PPP N (t)
with a rate of λ = 10 packets per minute. N (t) has the following probability function:
e−10t (10t)n
Pr [N (t) = n] = , n ∈ N.
n!
The probability of observing ten packets in a minute is
e−10 1010
Pr [N (1) = 10] = ≈ 0.125.
10!
Extended for the number of events in the interval (a + c, b + c], we have
Pr [N (a + c, b + c] = n] = Pr [N (a, b] = n]. (9.9)
From (9.9), it can be observed that the distribution of N (a + c, b + c] is independent
of c, while it only depends on the interval length, i.e., b − a. This accordingly verifies
the memoryless property of the SPPP.
In order to model the waiting lines in queuing systems with fixed arrival rate, the
SPPP is widely used since it approximately fits well in the actual behavior of these
N(t)
1
0 t
t1 t2 t3 t4 t5 t6
Figure 9.1 Interarrival time in Poisson point process N (t)
systems in practice. Specifically, in communication networks, the arrival of data

packets is random over time with a relatively short duration, which can be described
by the interarrival time modeled by an exponential random variable and the num-
ber of arrivals in a given time interval represented by a Poisson random variable
(cf. Figure 9.1). Due to the stationary, memoryless, and independent property of the
SPPP, the packet arrivals in a time interval have the same rate and have no influence on
the arrivals in other nonoverlapping time intervals. It can be noticed that this scenario
likely occurs when the data packets are generated from a large number of independent
sources and the characteristics of the interarrival distribution are invariant.
Theorem 9.1. In the SPPP, the time difference between events, a.k.a. interarrival
time, is exponentially distributed and independent with a mean of 1/λ.
Proof. Let us consider the interarrival time t1 , t2 , . . . , of a Poisson process, where ti

is the elapsed time between the ith arrival and the (i + 1)th arrival.
The first arrival occurs after time t if and only if there is no arrival in the interval
[0, t], i.e.,
Pr [t1 > t] = Pr [N (t) = 0] = e−λt . (9.10)
In general, the (i + 1)th arrival occurs after time ti + t if and only if there is no arrival
in the interval (ti , ti + t]. From (9.6), we have
Pr [ti+1 − ti > t] = Pr [N (ti , ti + t] = 0] = e−λt , (9.11)
and thus
Pr [ti+1 − ti ≤ t] = 1 − e−λt . (9.12)
Equation (9.12) is indeed the cumulative distribution function (cdf) of the expo-
nentially random variable with a mean of 1/λ. This means the interarrival time is
exponential distributed and independent. The lemma is proved.
This invariant average interval implies that the SPPP is stationary and memory-
less.
Adapting to queuing systems with a service facility, the next customer is served
as soon as the one in service leaves the system. The time between service completions
is indeed the service time. Therefore, if the time between completions is exponentially
distributed, then the service time itself is exponentially distributed. This means that
the service is also an SPPP.
9.2.2 Nonstationary Poisson point process

Definition 9.4. A counting process {N (t), t ≥ 0} is an inhomogeneous or NPPP char-
acterized by a rate λ(t) which varies over time. N (t) has the following probability
function:
e−(t) ((t))n
Pr [N (t) = n] = , n ∈ N. (9.13)
n!
Here, (t) denotes the average number of arrivals over time t, which can be
determined by
t
(t) = E[N (t)] = λ(s)ds.
0
Due to the stationary increment, the number of events in the interval (a, b], a ≤ b,
denoted by N (a, b] follows Poisson distribution with
e−(a,b) ((a, b))n
Pr [N (a, b] = n] = , n ∈ N, (9.14)
n!
where (a, b) denotes the average number of events in (a, b] given by
b
(a, b) = E[N (a, b]] = λ(t)dt.
a
The number of events is also independent in any disjoint intervals, i.e., (a1 , b1 ], (a2 , b2 ],
. . . , (aK , bK ], and thus
Pr [{N (a1 , b1 ] = n1 , N (a2 , b2 ] = n2 , . . . , N (aK , bK ] = nK }]

K
e−(ai ,bi ) ((ai , bi ))n
= i
. (9.15)
i=1
ni !
Example 9.2. Suppose e-mail arrivals to a server follow a PPP having rates of 2 per
minute in the morning from 8 am to 12 pm, and then 0.5 per minute in the afternoon
from 12 to 4 pm.
Set t = 0 corresponds to 8 am, an NPPP N (t) over time [0, t] has the following
rate function:

2 if 0 ≤ t < 4,
λ(t) =
0.5 if 4 ≤ t < 8.
The probability distribution of the number of arrivals over time [0, t] is expressed by
e−(t) ((t))n
Pr [N (t) = n] = , n ∈ N,
n!
where the average number of arrivals by time t, i.e., (t), is given by
⎧ t
⎪
⎪
⎪
⎪
⎪ 2ds = 2t if 0 ≤ t < 4,
⎪
⎨
(t) = 4 0
t
⎪
⎪
⎪
⎪ t
⎪
⎪ 2ds + 0.5ds = + 6 if 4 ≤ t < 8.
⎩ 2
0 4
The average number of arrivals between 11 am and 2 pm is computed by

4 6
(3, 6) = 2ds + 0.5ds = 3.
3 4
The probability distribution of the number of arrivals between 11 am and 2 pm is thus

given by
e−(a,b) ((a, b))n e−3 3n
Pr [N (3, 6) = n] = = .
n! n!
9.2.3 Operations of Poisson point process

We can perform mathematical operations on the PPP to generate different PPPs. Some
of them are as follows:
● Superposition/Pooling: The union of a countable set of PPPs N1 (t), N2 (t), . . . ,
forms a pooled process which is also a PPP due to the fact that a distribution
resulting from the sum of Poisson distributions retains the Poisson distribution.
Let N (t) denote the union set, i.e.,

N (t) = Ni (t). (9.16)
i
N (t) is a PPP with a mean of

(t) = i (t), (9.17)

i
where i (t) is the average number of arrivals over time t given by (9.4) for the
case of NPPP. The superposition property of the PPP means that if the arrivals to
a multi-server facility are PPP with each server having exponential service times,
the departures also constitute a Poisson stream.
● Splitting/Thinning: A PPP with a mean (t) can be split into multiple subpro-
cesses, each having a probability pi , i = 1, 2, . . . . The ith subprocess is also a
PPP with a mean rate of
i (t) = pi (t). (9.18)
Data collection
e
Probability distribution
identification
Parameter estimation
Distribution fitness test
Figure 9.2 Input modeling in network simulation
9.3 Input modeling

In any simulation model, the initial step is always to determine what kind of input
models should be suitable for the simulation. The development of an input model is
of great importance for building an appropriate simulation model.
Basically, there are four steps for developing an input model (cf. Figure 9.2),
which are as follows:
● Step 1 (data collection): to collect data from the real system.
● Step 2 (probability distribution identification): to identify a probability distribu-
tion to represent the input process given the collected data.
● Step 3 (parameter estimation): to estimate parameters for the distribution.
● Step 4 (distribution fitness test): to evaluate the selected distribution and
parameters via goodness-of-fit tests.
Details of the above steps will be discussed in the following subsections.
9.3.1 Data collection
Data collection is crucial in the development of an input model. The data is collected
to solve a variety of problems in a real system. The data collection process is based
on the principle of garbage-in–garbage-out, which is used to describe the nonsense
data output, i.e., “garbage,” caused by faulty data collection.
The data collection can be enhanced and facilitated by employing the following:
● Plan ahead: to predict circumstances via either practice or pre-observing session,
especially keeping track of unusual ones.
● Analyze data promptly: to evaluate the data as long as it is observed to check for
its appropriateness and accuracy.
● Combine data: to combine data sets of the same types over successive time
periods.
● Examine data: to perform data censoring over the entire collected data set.
● Check variables: to validate the relationship between variables in addition to
autocorrelation checking.
9.3.2 Probability distribution identification

Given the collected data, this step identifies the probability distribution of the data
set by developing a histogram of the data. The histogram represents the frequency
distribution of the data, through which the shape of the distribution can be determined.
In order to have an appropriate shape, the data distribution depends on the sample
size, i.e., the number of observations, and also the dispersion of the data.
Depending on data types, the histogram can be verified through the theoretical
distribution as follows:
● Continuous data: The shape of histogram should match the probability density
function.
● Discrete data: The shape of histogram should match the probability mass function.
Based on the shape of the histogram and the context of the input variables,
the distribution for random variables can be determined. Some well-known family
of distributions will be discussed in detail in Chapters 12 and 13 for discrete and
continuous random variables, respectively. Specifically, the physical basis of the
distribution can be exploited as a guide to select the appropriate distribution or an
approximated distribution. For instance, the following distributions can be identified
through their properties and functionalities:
● Binomial distribution: to represent the number of successes in a specified number
of trials.
● Poisson distribution: to represent the number of events that occur in a specified
amount of time or space.
● Normal distribution: to represent the sum of a number of processes of various
components.
● Exponential distribution: to represent the time between independent/memoryless
events.
● Weibull distribution: to represent the time to failure of components.
● Discrete or continuous uniform distribution: to represent equally likely outcomes
within a specified range.
● Triangular distribution: to represent a random process knowing only its minimum,
most likely and maximum values.
● Empirical distribution: to sample the observed data.
Note that the goal of the distribution identification step to find a good approx-
imation of the distribution function given the physical characteristics of the random
process. In order to evaluate the fitness of a distribution, a quantile–quantile (Q–Q)

plot can be exploited. The Q–Q plot is a graphical method which allows us to compare
two probability distributions through the shape of their quantiles. Here, the quantiles
are defined as the value sets of a variable which divide its histogram into contiguous
intervals with equal probabilities.
The quantile of a random variable can be defined by Definition 9.5.
Definition 9.5. If X is a random variable having cdf FX (x), then the q-quantile,
q ∈ [0, 1], of X is the α such that
FX (α) Pr [X ≤ α] = q. (9.19)
If FX (x) is invertible, then α can be determined as
α = FX−1 (q). (9.20)
Let xi , i = 1, 2, . . . , n, denote a sample of data from the random variable X and
yj , and j = 1, 2, . . . , n, denote the observations of the real data in ascending order. It
can be approximated that

−1 j − 0.5
yj ≈ F X (9.21)
n
where j is called the ranking or order number. This accordingly means that if X is
an appropriate
random variable
to represent the observed data, then the plot of { yj }
versus FX−1 (( j − 0.5)/n) is approximately a straight line. A further observation is
that if the line has a slope of one then the parameter values of the distribution are
appropriate.
Besides the check for appropriateness of a distribution, the Q–Q plot can be
employed to check the homogeneity between two sample sets by comparing their
plots of the order values. If they are homogeneous, then a distribution can represent
both of them.
9.3.3 Parameter estimation

The next step after selecting an appropriate family of distribution is to estimate param-
eters of the distributions. Specifically, mean and variance are two parameters that are
crucial in any distribution. Given a number of observed data, a.k.a. samples, their
sample mean and variance are defined by Definition 9.7.
Definition 9.6. Given a sampled data set of size N , i.e., {X1 , X2 , . . . , XN }, its sample
mean and variance are determined by
N
Xi
X̄ i=1 , (9.22)
N
N
X 2 − N X̄ 2
Var[X ] i=1 i . (9.23)
N −1
Definition 9.7. If the sampled data set are grouped with a frequency distribution fi
for the observed value Xi , the sample mean and variance can be given by
G
f i Xi
X̄ = i=1
G
, (9.24)
i=1 fi
G G
i=1 fi Xi −
2 2
i=1 fi X̄
Var[X ] = G , (9.25)
i=1 fi −1
where G is the number of groups in the data set.
Through statistical analysis over a number of observed values, the sample mean
and variance are regarded as two statistics of the distribution.
9.3.4 Distribution fitness test

The last step in the development of an input model is to evaluate the selected distribu-
tion and the estimated parameters. The appropriateness of the input data distribution
can be examined by conducting hypothesis testing, a.k.a. goodness-of-fit test, which
is defined as a statistical model used to determine how a distribution fits a sampled
data set. Some goodness-of-fit tests can be employed, such as Kolmogorov–Smirnov
(K–S) test, chi-squared test for large sample size, Fisher test for small sample size,
etc. Note that there is no single correct distribution in a real application, especially
when there are numerous data samples available.
Intuitively, a hypothesis test can be carried out by comparing the histogram of
the data samples with the shape of the probability mass function for discrete data
or the probability density function for continuous data. Specifically, the K–S test is
based on the comparison of the distance between the empirical distribution function
of the sample and the cdf of the reference distribution to evaluate the fitness of the
candidate distribution.
9.4 Related works
references:
Statistical Models Davison in 2003 [210], Cox in 2006 [211], Rissanen in 2007
[212], Wasserman in 2010 [213], and Kroese and Chan in 2013 [214].
Probability Distribution Papoulis in 2002 [215], Forbes et al. in 2010 [216],
Montgomery and Runger in 2013 [217], and Ross in 2014 [218].
Poisson Point Process Ross in 1995 [219], Nelson in 1995 [220], Kulkarni in 1995
[202], Papoulis in 2002 [215], Daley and Vere-Jones in 2003 and 2008 [221,222],
Illian et al. in 2008 [223], Bass in 2011 [224], and Snyder and Miller in 2012
[225].
Input modeling Jain in 1991 [114], Banks et al. in 2010 [185], Choi and Kang in
2013 [205], and Obaidat in 2015 [226].
Parameter Estimation Bos in 2007 [227], Levy in 2008 [228], and Aster et al. in
2012 [229].
Distribution Fitness Test Massey in 1951 [230], D’Agostino in 1986 [231], Green-
wood and Nikulin in 1996 [232], Fan in 1997 [233], and Justel et al. in 1997
[234].
[Q9.1] Provide some statistical models which are commonly used in queuing
systems.
[Q9.2] How to model packet interarrival time, service time, and failure time in a
queuing model?
[Q9.3] What is a Poisson point process (PPP)?
[Q9.4] What is the probability of observing a number of events x within a specified
interval in a PPP having an event rate of λ?
[Q9.5] Differentiate between stationary PPP and nonstationary PPP.
[Q9.6] What is the probability function of stationary PPP and nonstationary PPP
within a specific interval t?
[Q9.7] Explain how to generate different PPPs with superposition/pooling and
splitting/thinning operations.
[Q9.8] Describe the steps for developing an input model in network simulation.
Illustrate them with a flowchart.
[Q9.9] How to verify the appropriateness of the collected data?
[Q9.10] Explain how to identify a probability distribution of the collected data.
[Q9.11] Provide some typical distributions that can be easily identified through
their properties.
[Q9.12] What are quantiles of a random variable?
[Q9.13] How to determine the q-quantile of a random variable X ?
[Q9.14] Explain how to employ Q–Q plot to evaluate the fitness of a distribution
of a random variable.
[Q9.15] What are sample mean and sample variance of a data set and how to
determine them?
[Q9.16] What is distribution fitness test and its principle in examining a data
distribution?
9.6 Problems
[P9.1] Show that the interarrival time in both stationary and nonstationary PPP
follows an exponential distribution.
[P9.2] Does the probability function increase or decrease if the rate of the PPP
increases? Prove this.
[P9.3] Consider packet arrivals at a server following a PPP with a rate of 25 packets
per minute.
● Derive the probability function of the PPP.
● Find the probability that there are 30 packet arrivals in a minute.
● Assume the rate of the PPP is doubled, i.e., 50 packets per minute, find
the probability that there are 30 packet arrivals in a minute.
[P9.4] E-mails arrive at a server following a PPP with a rate of 1,000 e-mails per
day during the weekday, i.e., from Monday to Friday, and 200 e-mails per
day during the weekend. Assume a week starts on Monday, i.e., t = 0, and
ends on Sunday, i.e., t = 6.
● Derive the probability function of the number of e-mails over time [0, t]
where 0 ≤ t ≤ 6.
● Derive the probability function of the number of e-mails received from
Wednesday to Saturday.
[P9.5] Provide two examples to illustrate the pooling and thinning of PPP.
[P9.6] Provide an example to show that Q–Q plot can be used to check the
homogeneity between two sample sets.
[P9.7] Do the research on goodness-of-fit tests. Discuss the principle and usage of
Kolmogorov–Smirnov test and Fisher test for various sample sets.
Chapter 10
Probabilities in performance analysis
When talking about the randomness of an event in statistical models, it always goes
together with the probability concept. This chapter is devoted to introduce the prob-
abilities in the context of network performance analysis. For some people who may
have learned and are familiar with the general concept of the probability, this chap-
ter will help them review these and present the terminology and concepts in the
communication networks.
In this chapter, I will first introduce the basic concepts and terminology of the
probability in Section 10.1. Section 10.2 will present axioms and properties of the
probability. Considering the relationship between events, Sections 10.3 and 10.4 will
discuss the conditional probability and independence of the events along with their
applications in network-performance evaluation.
10.1 Basic concepts and terminology of probability

In statistical theory, some well-known concepts and terminology of the probability
can be defined as follows:
Definition 10.1. An experiment is an examination or a trial run conducted through

a procedure of finding, an observation of fact(s) and a model of data.
Example 10.1. An experiment could consist of the following:

● A procedure of finding a server for data storage among a number of available
servers.
● An observation of which servers are appropriate or inappropriate.
● A model of selecting an appropriate server after a specified number of trials given
various constraints on the criteria of server selection.
Definition 10.2. An outcome is any possible observation of an experiment.
Definition 10.3. A sample space is a set of all possible outcomes of an experiment

satisfying that the set is finest grain, mutually exclusive, and collectively exhaustive.
In order to understand clearly Definition 10.3, in-depth knowledge of set the-

ory is required. The concept “finest grain,” “mutually exclusive,” and “collectively
exhaustive” can be simply understood as follows:
● Finest grain set: is a set consisting of all distinguishable outcomes.
● Mutually exclusive set: A set S of N elements xi , i = 1, 2, . . . , N , is mutually
exclusive when
xi = xj , i = j. (10.1)
If the set S has K subsets S1 , S2 , . . . , SK ⊂ S, then S is a mutually exclusive
set when
Si ∩ Sj = ∅, i = j, (10.2)
where ∅ denotes a null set which has no elements and ∩ is an intersection
operation of sets corresponding to logical AND operation (denoted by ∧). By
definition,
x ∈ A ∩ B ⇔ x ∈ A ∧ x ∈ B. (10.3)
● Collectively exhaustive set: is a set consisting of every outcomes, i.e.,

K
S = S1 ∪ S2 ∪ · · · ∪ SK Si , (10.4)
i=1
where ∪ is a union operation of sets corresponding to logical OR operation

(denoted by ∨). By definition,
x ∈ A ∪ B ⇔ x ∈ A ∨ x ∈ B. (10.5)
Remark 10.1. A set is mutually exclusive and collectively exhaustive when it satisfies
both (10.2) and (10.4).
Depending on data types, a sample space can be either a discrete or continuous set.
Example 10.2. The sample space when decoding three received data packets is
S = {s1 s2 s3 , s1 s2 f3 , s1 f2 s3 , f1 s2 s3 , s1 f2 f3 , f1 s2 f3 , f1 f2 s3 , f1 f2 f3 },
where si and fi , i = 1, 2, 3, are the successful and failed outcomes, respectively, of the
ith packet.
Definition 10.4. An event is a specified set/subset of outcomes in an experiment. The
corresponding event space is a mutually exclusive and collectively exhaustive set of
all the outcomes of the event.
An illustration of event in an experiment is shown in Figure 10.1 where the sample
space of the experiment has eight outcomes, and a specified event consisting of two
outcomes is selected.
Example 10.3. Suppose we examine three decoded data packets. Let si and fi ,
i = 1, 2, 3, denote the successful packet and failed packet, respectively, of the ith
Probabilities in performance analysis 161
Outcome
Outcome
Outcome
Outcome Outcome
Outcome
Event
Outcome
Outcome
Sample
space
Experiment
Figure 10.1 Event in an experiment
packet. An event E consists of all possible outcomes which have at least two successful
packets. The corresponding event space E of the event E is
E = {s1 s2 s3 , s1 s2 f3 , s1 f2 s3 , f1 s2 s3 },
Definition 10.5. An experiment is random if its outcome cannot be predicted with
certainty or determined with a probability.
Most experiments in practice are random. For instance, the communication chan-
nels in wireless networks suffer from random fading and noise due to the randomness
of the communication media.
10.2 Axioms and properties of probability

In a sample space, each event is mapped to a real number called a probability mea-
sure. The probability of an event E, denoted by Pr [E], also means the likelihood of
outcomes in that event.
Remark 10.2. Two well-known axioms of the probability are

0 ≤ Pr [E] ≤ 1 (10.6)

K
Pr [S] = Pr [Si ] = 1, (10.7)
i=1
K
where S is a sample space having S = i=1 Si and Si ∩ Sj = ∅∀i = j.
Equation (10.7) can also be applied for any sets having a number of subsets as
in Theorem 10.1.
Theorem 10.1. If a set A has M collectively exhaustive and mutually exclusive subsets
{A1 , A2 , . . . , AM }, then
M

M
Pr [A] = Pr Ai = Pr [Ai ]. (10.8)
i=1 i=1
Proof. The proof follows from the definition of union and intersection operations of
collectively exhaustive set and mutually exclusive set, respectively, with A = A1 ∪
A2 ∪ · · · ∪ AM and Ai ∩ Aj = ∅, ∀i = j.
A special case of Theorem 10.1 is when considering an event with a specified

number of outcomes as in Lemma 10.1 and Corollary 10.1.
Lemma 10.1. If an event E has M outcomes, i.e., E = {e1 , e2 , . . . , eM }, then

M
Pr [E] = Pr [ei ]. (10.9)
i=1
Proof. The proof follows from Theorem 10.1 where each outcome ei , i = 1, 2, . . . , M ,
is regarded as a subset of a set E.
Corollary 10.1. If each outcome ei , i = 1, 2, . . . , M , of a sample space S is equally

likely, i.e., Pr [e1 ] = Pr [e2 ] = · · · = Pr [eM ], then
1
Pr [ei ] = . (10.10)
M
Proof. The proof follows

from the probability axiom Pr [S] = 1 and from Lemma
10.1 with Pr [S] = M i=1 Pr [ei ] = M Pr [ei ].
Example 10.4. A data symbol consists of 2 b. There are four different data symbols
which can be generated from these 2 b. Assume that they are equally likely. Applying
Corollary 10.1, the probability of each outcome is 1/4.
Some other properties of probability can be obtained from the set theory concepts
as in Theorems 10.2 and 10.3.
Theorem 10.2. A null set ∅ has

Pr [∅] = 0. (10.11)
Proof. The proof follows from the fact that there is nothing in the null set.
Theorem 10.3. A complement of set A, denoted by A, has

Pr [A] = 1 − Pr [A]. (10.12)
Proof. The proof follows from the definition of the complement set, i.e.,
x ∈ A ⇔ x ∈ A. (10.13)
Equivalently, we have
Pr [A] + Pr [A] = 1. (10.14)
Example 10.5. A data packet is received correctly with a probability of 0.2. Then,
the probability of the packet to be corrupted or lost is 1 − 0.2 = 0.8.
Theorem 10.4. A subset Ai of a set A has

Pr [Ai ] ≤ Pr [A]. (10.15)
Proof. The proof follows from the definition of a subset, i.e., if x ∈ Ai , then
x ∈ A.
Example 10.6. A data symbol consists of 2 b 0 and 1. Four different data symbols
generated from these 2 b are assumed to be equally likely. Let A denote the event that
the data symbol consists of at least 1 b; 1 and B denote the event that the data symbol
consists of exact 2 b 1. Events A and B can have the following possible outcomes:
A = {01, 10, 11}
B = {11}
It can be seen that B ⊂ A, Pr [B] = 1/4, and Pr [A] = 3/4. This verifies that Pr [B] ≤
Pr [A] as shown in Theorem 10.4.
Theorem 10.5. Given any two sets A and B, we have

Pr [A ∪ B] = Pr [A] + Pr [B] − Pr [A ∩ B] (10.16)
Proof. The proof follows from the definition of union set. The union set of A and
B is equal to set A plus set B minus the intersection set between A and B, i.e.,
A ∩ B.
Note that the probability of intersection set A ∩ B can be simply written as

Pr [AB].
Example 10.7. Suppose that a source S wants to send data packets to both destina-
tions D1 and D2 . The probability that the packet is received correctly: (i) at D1 is 0.3,
(ii) at D2 is 0.4, and (iii) at both D1 and D2 is 0.1. Then, by employing (10.16) in
Theorem 10.5, the probability that the packet is received correctly at either D1 , D2 ,
or both can be found as 0.3 + 0.4 − 0.1 = 0.6.
Theorem 10.6. Given probabilities of mutually exclusive intersection sets A ∩

B1 , A ∩ B2 , . . . , A ∩ BM , the probability of set A can be determined by

M
Pr [A] = Pr [ABi ]. (10.17)
i=1
Proof. The proof follows from the fact of union and intersection sets that

M
A = (A ∩ B1 ) ∪ (A ∩ B2 ) ∪ · · · ∪ (A ∩ BM ) = (A ∩ Bi ), (10.18)
i=1
and from the computation of the probability of the union set as in Theorem 10.1.
Example 10.8. A model of mobile phone usage consists of voice ( V ) and data (D).
Each voice or data call can be classified as either long (L) or short (S). The probability
of a long voice call, i.e., Pr [LV ], is 0.2, of a short voice call, i.e., Pr [SV ], is 0.3, of
a long data call, i.e., Pr [LD], is 0.4, and of a short data call, i.e., Pr [SD], is 0.1.
Applying Theorem 10.6, we can obtain the probability of a long call, i.e., Pr [L], and
of a voice call, i.e., Pr [V ], as follows:
Pr [L] = Pr [LV ] + Pr [LD] = 0.2 + 0.4 = 0.6
Pr [V ] = Pr [LV ] + Pr [SV ] = 0.2 + 0.3 = 0.5.
10.3 Conditional probability
In a practical system, events generally occur in sequence. The probability that an

event occurs given that another event has occurred is called a conditional probability.
This section will introduce the concept of conditional probability.
Definition 10.6. The conditional probability of an event A given an event B is denoted
by Pr [A|B] which can be determined by
Pr [AB]
Pr [A|B] . (10.19)
Pr [B]
Remark 10.3. Definition 10.6 also means that in order for both events A and B
to occur, event B can occur first and then event A occurs given that B has already
occurred. Equation (10.19) can be equivalently written as
Pr [AB] = Pr [A|B]Pr [B]. (10.20)
Remark 10.4. A notice in (10.19) is that there is an implicit condition of Pr [B] > 0.
Indeed, if Pr [B] = 0, i.e., event B does not exist, then Pr [A|B] has no meaning.
Example 10.9. The validation of a communication channel between two devices is

performed with two error checks. 20% of the data packets have errors that can be
detected in both checks, while 40% of the data packets can only be detected for errors
in the first check.
Let A and B denote the events that the packets can be detected for errors in the first
check and in the second check, respectively. We have Pr [AB] = 0.2 and Pr [A] = 0.4.
The probability of the packets that can be detected for errors in the second check
if they can also be detected in the first check is
Pr [AB] 0.2
Pr [B|A] = = = 0.5.
Pr [A] 0.4
Remark 10.5. Similar to the axioms of the probability, we have some well-known
axioms of the conditional probability as follows:
0 ≤ Pr [A|B] ≤ 1, (10.21)
Pr [A|A] = 1. (10.22)
Some other properties of the conditional probability are shown in the following
theorems:
Theorem 10.7. If a set A has M collectively exhaustive and mutually exclusive subsets
{A1 , A2 , . . . , AM }, then
M

M
Pr [A|B] = Pr Ai |B = Pr [Ai |B]. (10.23)
i=1 i=1
Proof. The proof follows from Definition 10.6 and Theorem 10.1, i.e.,
Pr [AB]
Pr [A|B] =
Pr [B]

M
Pr i=1 Ai B
=
Pr [B]
M

= Pr Ai |B
i=1

M
= Pr [Ai |B].
i=1
Theorem 10.8 (Bayes’ theorem). The relationship between conditional probabilities

of two events A and B is
Pr [B|A] Pr [A]
Pr [A|B] = . (10.24)
Pr [B]
Proof. From Definition 10.6 of the conditional probability, Pr [B|A] is given by

Pr [AB]
Pr [B|A] =
Pr [A]
or
Pr [AB] = Pr [B|A] Pr [A]. (10.25)
Substituting (10.25) into (10.19) completes the proof.
Example 10.10. A server carries out two tests on every received data packets. 25%
of the packets pass both tests, while 40% of the packets pass the first test, and 30%
of the packets pass the second test.
Let A and B denote the event that the packets pass the first test and the second
test, respectively. We have Pr [AB] = 0.25, Pr [A] = 0.4, and Pr [B] = 0.3
The probability of the packets that pass the second test if they also pass the first
test is
Pr [AB] 0.25
Pr [B|A] = = = 0.625.
Pr [A] 0.4
The probability of the packets that pass the first test if they also pass the second
test is
Pr [AB] 0.25
Pr [A|B] = = = 0.833.
Pr [B] 0.3
Pr [A|B] can also be interpreted as the probability that the accepted data packets in
the second test can pass the first test.
Remark 10.6. It can be noted in example that Pr [A|B] = Pr [B|A]. These two
conditional probabilities also have different meanings.
A special case is when two events A and B are equally likely as in the following
lemma.
Lemma 10.2. If two events A and B are equally likely, then

Pr [A|B] = Pr [B|A]. (10.26)
Proof. The proof follows from Theorem 10.8 when Pr [A] = Pr [B].
Considering the intersection of different events, we have a law of total probability

in Theorem 10.9.
Theorem 10.9. Given probabilities of M events {B1 , B2 , . . . , BM } and conditional

probabilities of event A given Bi , i = 1, 2, . . . , M , the probability of event A can be
determined by

M
Pr [A] = Pr [A|Bi ] Pr [Bi ]. (10.27)
i=1
Proof. The proof follows from Theorem 10.6 of the intersection sets and Definition
10.6 of conditional probability

M
Pr [A] = Pr [ABi ]
i=1

M
= Pr [A|Bi ] Pr [Bi ].
i=1
Example 10.11. A cloud computing data center has three servers, denoted by S1 , S2 ,
and S3 , to process the data. 20% of the data is processed at the first server, 30% of
data at the second server, and 50% of the data at the third server.
Let A denote the event that the data is successfully processed at the data center.
Observing the successful data process from the three servers, we have the follow-
ing conditional probabilities: Pr [A|S1 ] = 0.8, Pr [A|S2 ] = 0.7, and Pr [A|S3 ] = 0.9.
The probability that the data is successfully processed can be determined by the
law of total probability in Theorem 10.9 as
Pr [A] = Pr [A|S1 ] Pr [S1 ] + Pr [A|S2 ] Pr [S2 ] + Pr [A|S3 ] Pr [S3 ]
= 0.8 × 0.2 + 0.7 × 0.3 + 0.9 × 0.5 = 0.82.

By applying the Bayes’ theorem (cf. Theorem 10.8), the probabilities that the
successfully processed data comes from the first, second, and third server are,
respectively, given by
Pr [A|S1 ] Pr [S1 ] 0.8 × 0.2
Pr [S1 |A] = = ≈ 0.195,
Pr [A] 0.82
Pr [A|S2 ] Pr [S2 ] 0.7 × 0.3

Pr [S2 |A] = = ≈ 0.256,
Pr [A] 0.82
Pr [A|S3 ] Pr [S3 ] 0.9 × 0.5

Pr [S3 |A] = = ≈ 0.549.
Pr [A] 0.82
It can be noticed that
Pr [S1 |A] + Pr [S2 |A] + Pr [S3 |A] = 1.
This accordingly means that the successfully processed data has to be from either one
of three servers, and this is always true. However,
Pr [A|S1 ] + Pr [A|S2 ] + Pr [A|S3 ] = 1.
From the observation in Example 10.11, we have the following lemma.
Lemma 10.3. Given M events {B1 , B2 , . . . , BM }, the conditional probabilities of event

Bi , i = 1, 2, . . . , M , given an event A should satisfy

M
Pr [Bi |A] = 1. (10.28)
i=1
Proof. The proof follows from the law of total probability in Theorem 10.9 and Bayes’
theorem, i.e., Theorem 10.8, as

M
M
Pr [A|Bi ] Pr [Bi ]
Pr [Bi |A] =
i=1 i=1
Pr [A]
1
M
= Pr [A|Bi ] Pr [Bi ]
Pr [A] i=1
1
= Pr [A] = 1.
Pr [A]
10.4 Independence of events

Events in practical systems can be either dependent or independent of each other. The
independence of two events can be defined as follows:
Definition 10.7. Two events A and B are independent if and only if the probability of
their intersection set equals to the product of their individual probabilities, i.e.,
Pr [A ∩ B] = Pr [AB] = Pr [A] Pr [B]. (10.29)
Example 10.12. Consider two independent checks of data packets arriving at a

server. The acceptance rates in the first check and the second check are 0.5 and 0.6,
respectively.
Let A and B denote the events that the packets pass the first check and the second
check, respectively. We have Pr [A] = 0.5 and Pr [B] = 0.6.
Since A and B are independent, the probability that the packets pass both checks is
Pr [AB] = Pr [A] Pr [B] = 0.5 × 0.6 = 0.3.
Intuitively, if two events A and B are independent, then the occurrence of them
does not affect each other. This can be shown in Theorem 10.10.
Theorem 10.10. If two events A and B are independent, then

Pr [A] = Pr [A|B], (10.30)
Pr [B] = Pr [B|A]. (10.31)
Proof. The proof follows from Definition 10.6 of the conditional probability and
Definition 10.7 of the independent events, i.e.,
Pr [AB]
Pr [B] =
Pr [A|B]
Pr [A] Pr [B]
= .
Pr [A|B]
We thus deduce (10.30) and similarly obtain (10.31), which completes the proof.
The above definition of two independent events can be extended to the general
case of more than two events as follows:
Definition 10.8. M events {A1 , A2 , . . . , AM } are independent if and only if every set
of (M − 1) sets is independent and
M

M
Pr Ai = Pr [Ai ]. (10.32)
i=1 i=1
Considering conditionally independent events, we have the following property:
Theorem 10.11. If two events A and B are conditionally independent given an event
C, then
Pr [AB|C] = Pr [A|C] Pr [B|C] (10.33)
and
Pr [A|BC] = Pr [A|C]. (10.34)
Proof. The proof of (10.33) follows from the Bayes’ theorem (cf. Theorem 10.8) and
Definition 10.6 as
Pr [AC] Pr [BC]
Pr [AB|C] =
Pr [C] Pr [C]
= Pr [A|C] Pr [B|C]
The proof of (10.34) follows from Theorem 10.10, i.e.,
Pr [A|BC] = Pr [A|B|C]
= Pr [A|C].
Remark 10.7. The independence concept is also related to the RMA analysis in
Chapter 5 (cf. Section 5.9).
Let S denote the event that the whole system consisting of N independent devices
is successful and Dn , n = 1, 2, . . . , N , denote the event that the nth device is successful.
If these devices are connected in series, it can be deduced that
N

N
Pr [S] = Pr Dn = Pr [Dn ]. (10.35)
n=1 n=1
This accordingly verifies (5.13). Similarly, if N independent devices are connected in

parallel, we can arrive at (5.16), i.e.,
N
Pr [S] = 1 − (1 − Pr [Dn ]). (10.36)
n=1
10.5 Related works

references: Feller in 1968 [235], Nelson in 1995 [220], Papoulis in 2002 [215], Yates
and Goodman in [236], Gubner in 2006 [237], Leon-Garcia in 2007 [238], Durrett
in 2010 [239], Montgomery and Runger in 2013 [217], Ross in 2014 [218], Gallager
in 2014 [240], Steyer and Nagel in 2017 [241], and Mitzenmacher and Upfal in
2017 [242].

[Q10.1] What is an experiment? Give an example of an experiment in wireless
communications and explain its possible outcomes.
[Q10.2] How to determine sample space of an experiment?
[Q10.3] Differentiate between mutually exclusive set and collectively exhaustive
set.
[Q10.4] What is an event in an experiment? Give an example of an event in the
experiment in Question [Q10.1].
[Q10.5] What is the probability of an event having M outcomes each of which
occurs with a probability of Pi , i = 1, 2, . . . , M ?
[Q10.6] If all M outcomes of a sample space of an experiment are equally likely,
what is the probability of each outcome?
[Q10.7] What is a complement set and how to determine its probability?
[Q10.8] How to compute the probability of a union of two sets A and B?
[Q10.9] Determine the probability of a set A given the probabilities of mutually
exclusive intersection sets A ∩ B1 , A ∩ B2 , . . . , A ∩ BM .
[Q10.10] What is the conditional probability of an event A given an event B?
[Q10.11] Describe Bayes’ theorem and prove it.

[Q10.12] Determine the probability of an event A given the probabilities of M events
and the conditional probabilities of A given each of these M events.
[Q10.13] Prove that the summation of all conditional probabilities of event Bi ,
i = 1, 2, . . . , M , given an event A is always one.
[Q10.14] What is the condition of two independent events?
10.7 Problems
[P10.1] Monitor three consecutive phone calls going through a telephone switch-
ing office. Classify each one as a voice call (v) if someone is speaking or a
data call (d) if the call is carrying a modem or fax signal. Your observation
is a sequence of three letters (each letter is either v or d). For example,
two voice calls followed by one data call corresponds to vvd. Write the
elements of the following sets:
● A1 = {first call is a voice call}
● A2 = {second call is a voice call}
● A3 = {all calls are the same}
● A4 = {one or more voice calls}
● B1 = {first call is a data call}
● B2 = {second call is a data call}
● B3 = {voice and data alternate}
● B4 = {two or more data calls}
For each pair of events A1 and B1 , A2 and B2 , and so on, identify whether
the pair of events is either mutually exclusive or collectively exhaustive
or both.
[P10.2] Differentiate between mutually exclusive events and independent events.
[P10.3] Derive the probability of the union of two events A and B for the following
cases:
● A and B are mutually exclusive.
● A and B are independent.
[P10.4] Monitor a phone call. Classify the call as a voice call (V ) if someone
is speaking or a data call (D) if the call is carrying a modem or fax
signal. Classify the call as long (L) if the call lasts for more than 3 min;
otherwise, classify the call as brief (B). Based on data collected by the
telephone company, we use the following probability model: P[V ] = 0.7,
P[L] = 0.6, P[VL] = 0.35. Find the following probabilities:
● Pr [DL]
● Pr [D ∪ L]
● Pr [VB]
● Pr [V ∪ L]
● Pr [V ∪ D]
● Pr [LB]
[P10.5] Monitor three consecutive phone calls going through a telephone switch-
ing office. Classify each one as a voice call (v) if someone is speaking or
a data call (d) if the call is carrying a modem or fax signal. Your observa-
tion is a sequence of three letters (each one is either v or d). For example,
three voice calls corresponds to vvv. The outcomes vvv and ddd have
probability 0.2, whereas each of the other outcomes vvd, vdv, vdd, dvv,
dvd, and ddv has probability 0.1. Count the number of voice calls NV
in the three calls you have observed. Consider the four events NV = 0,
NV = 1, NV = 2, NV = 3. Describe in words and calculate the following
probabilities:
● Pr [NV = 2]
● Pr [NV ≥ 1]
● Pr [{vvd}|NV = 2]
● Pr [{ddv|}|NV = 2]
● Pr [NV = 2|NV ≥ 1]
● Pr [NV ≥ 1|NV = 2]
Determine whether the pair of events {NV = 2} and {NV ≥ 1} is
independent.
[P10.6] In a cellular phone system, a mobile phone must be paged to receive a
phone call. However, paging attempts do not always succeed because the
mobile phone may not receive the paging signal clearly. Consequently, the
system will page a phone up to three times before giving up. If a single
paging attempt succeeds with probability 0.8, sketch a probability tree
for this experiment and find the probability P[F] that the phone is found.
[P10.7] A machine produces routers in pairs. Tests show that the first router is
acceptable with probability 70%. If the first router is acceptable, then the
second router will be acceptable with probability 90%. Otherwise, if the
first router is defective, then the second router will be acceptable with
probability 30%.
● Find the probability that exactly one router of a pair is acceptable.
● Find the probability that both routers in a pair are acceptable.
● Suppose the quality of each pair of routers is independent of the
quality of every other pair of routers. Find the probability of at least
one acceptable router out of two pairs of routers.
● Find the minimum number of pair of routers the machine must produce
to reach a probability of 99% that there will be at least one acceptable
router.
[P10.8] Consider a binary code with 4 b (0 or 1) in each code word. An example
of a code word is 0110.
● Find the number of code words that can be generated.
● Find the number of code words that have exactly two zeroes.
● Find the number of code words that begin with a zero.
● In a constant-ratio binary code, each code word has N bits. In every
word, M of the N bits are 1 and the other N − M bits are 0. Find the
number of code words in the code when N = 8 and M = 3.
[P10.9] Consider 64-quadrature amplitude modulation (64-QAM) with equally

likely data symbols. Let A and B denote the events that the data symbol
consists of at least 4 b 1 and at least 4 b 0, respectively.
● Find the number of bits in each 64-QAM symbol.
● List all possible data symbols in A and B.
● Find Pr [A] and Pr [B].
● Find Pr [A ∩ B] and Pr [A ∪ B].
● Verify if A and B are either mutually exclusive, collectively exhaustive,
or independent.
● Find Pr [A|B] and Pr [B|A].
[P10.10] A server performs two tests on N data packets. Suppose that M packets
pass both tests, M1 packets pass only the first test and M2 packets pass
only the second test. Let A and B denote the events that the packets pass
the first test and pass the second test, respectively.
● Find Pr [A ∩ B] and Pr [A ∪ B]
● Find Pr [A|B] and Pr [B|A].
● Find the conditions of N , M , M1 , and M2 that A and B are independent.
[P10.11] A data center has K servers to process the data. Assume load balancing
where each server processes the same amount of data. Observing the data
processing at each server, we notice the kth server, k = 1, 2, . . . , K, has
a success rate of pk .
● Find the probability that the data is successfully processed at the data
center.
● Find the probability that the successful processed data comes from
the kth server where k = 1, 2, . . . , K.
[P10.12] A memory module consists of M chips. The device is designed with
redundancy so that it works even if one of its chips is defective. Each
chip contains N transistors and functions properly if all of its transistors
work. A transistor works with a probability p independent of any other
transistor. Find the probability that a chip works and the probability that
a memory module works.
Chapter 11
Random variables in network modeling
and simulation
In network modeling and simulation, a real network can be represented via statistical
models. Due to the randomness of the phenomenon, possible outcomes of the event
can be assigned by a real number of random quantity which is known as a random
variable. The random variable is specified by a probability distribution. Depending on
the data type of the random quantity, we have two kinds of random variables including
discrete and continuous random variables.
In this chapter, I will first introduce the basic concepts of random variables in
Section 11.1. Discrete and continuous random variables will be sequentially presented
in Sections 11.2 and 11.3. Specifically, distribution functions, mean, variance, and
moments will be discussed for each random variable type.
11.1 Introduction of random variables

In an experiment, a probability measure is employed for every random event over a
sample space S. A random variable (cf. Figure 11.1) can be defined as follows:
Definition 11.1. A random variable X is a mapping function that assigns a real
number to each outcome of the experiment, i.e.,
X : S → R. (11.1)
where S is the sample space consisting of all possible outcomes in the experiment.
Let ω denote a sample point in S, then X (ω) ∈ R is the measured value of ω.
Example 11.1. Consider a device-to-device communication channel between two

mobile devices A and B. The device A wants to share two files with B. Due to the long
distance between them, this sharing can be either successful (s) or failed (f). The set
of all possible outcomes is S = {ss, sf , fs, ff }.
Let a random variable X represent the number of files that are successfully
received and ω denote a sample point in S. We have
⎧
⎪
⎨0 if ω = ff ,
X (ω) = 1 if ω = sf , fs,
⎪
⎩
2 if ω = ss.
Outcome Real value
Random
Outcome Real value variable
X
Outcome Real value

Random
event
Figure 11.1 Random variable for representing a random event
A random variable is represented by a probability value corresponding to the

probability of an event. For instance, the probability that X has a value of x is1 written
as Pr [X (ω) = x] or simply Pr [X = x]. Aggregating all probability values of a random
variable constructs a probability distribution of that random variable.
Basically, there are two types of random variables with respect to its data
type which could be either discrete or continuous random variables. These ran-
dom variables have different distribution functions. In the following sections, I will
discuss in detail these two types of random variables along with their distribution
functions.
11.2 Discrete random variables

Definition 11.2. A random variable X : S → R is a discrete random variable if the
range of X is a countably infinite set, i.e.,
SX = {x1 , x2 , . . . } (11.2)
or a finite set of N elements, i.e.,
SX = {x1 , x2 , . . . , xN }. (11.3)
1
Note that the upper case letter is used to name the random variable, while the lower case letter is for its
value.
Random variables in network modeling and simulation 177
11.2.1 Probability mass function

A discrete random variable is described by a probability mass function (pmf) which
can be defined as follows:
Definition 11.3. The pmf fX (x) of a discrete random variable X assigns a probability
to each measured value x of the outcome, i.e.,
fX (x) Pr [X = x]. (11.4)
0.5
0.45
0.4
0.35
0.3
fX (x)
0.25
0.2
0.15
0.1
0.05
0
0 1 2
x
Figure 11.2 pmf of random variable in Example 11.2
Example 11.2. In Example 11.1, it is assumed that the success and fail of file
sharing between two devices are equally likely. The outcomes in the sample space
S = {ss, sf , fs, ff } are thus equally likely, each of which has a probability of 1/4. The
pmf of the number of files that are successfully received, i.e., random variable X (cf.
Figure 11.2), is given by
⎧
⎪
⎨1/4 if x = 0, 2,
fX (x) = 1/2 if x = 1,
⎪
⎩
0 otherwise.
Remark 11.1. From Definition 11.3, it can be noticed that the pmf of a discrete
random variable X has the following properties:
0 ≤ fX (x) ≤ 1 ∀x ∈ SX , (11.5)

fX (x) = 1. (11.6)
x∈SX
Similar to conditional probability, we have conditional pmf which is defined as

follows:
Definition 11.4. The conditional pmf of a random variable X given an event B is
denoted by fX |B (x) which can be determined by
fX |B (x) Pr [X = x|B] (11.7)
Pr [X = x, B]
= . (11.8)
Pr [B]
We also have the law of total pmf as follows:
Theorem 11.1. Given probabilities of M events {B1 , B2 , . . . , BM } and conditional pmf

of a random variable X given Bi , i = 1, 2, . . . , M , the pmf of X can be determined by

M
fX (x) = fX |B (x) Pr [Bi ]. (11.9)
i=1
Proof. The proof is similar to that of Theorem 10.9.

Considering a derived random variable Y which is a function of X , i.e., Y = g(X ),
its pmf can be determined by the following theorem:
Theorem 11.2. The pmf of a derived discrete random variable Y = g(X ) is given by

fY ( y) = fX (g −1 (y)). (11.10)
x:x=g −1 (y)
Proof. The proof follows from Definition 11.3, i.e.,

fY ( y) = Pr [Y = y]
= Pr [g(X ) = y]

= Pr [X = g −1 (y)]
x:x=g −1 (y)

= fX (g −1 (y))
x:x=g −1 (y)
11.2.2 Cumulative distribution function

In addition to pmf, in order to represent a discrete random variable, there is a related
function, namely, cumulative distribution function (cdf) which is defined as follows:
Definition 11.5. The cdf FX (x) of a discrete random variable X is the probability
that X will take a value less than or equal to x, i.e.,
FX (x) Pr [X ≤ x] (11.11)

= fX (x ). (11.12)
x ≤x∈SX
0.9
0.8
0.7
0.6
0.5
FX (x)
0.4
0.3
0.2
0.1
0
–1 –0.5 0 0.5 1 1.5 2 2.5 3
x
Figure 11.3 cdf of random variable in Example 11.2
Example 11.3. In Example 11.1, the cdf of the number of files that are successfully
received, i.e., random variable X (cf. Figure 11.3), is given by
⎧
⎪
⎪ 0 if x < 0,
⎪
⎨1/4 if 0 ≤ x < 1,
FX (x) =
⎪3/4 if 1 ≤ x < 2,
⎪
⎪
⎩
1 if x ≥ 2.
Remark 11.2. From Definition 11.5, it can be noticed that the cdf of a discrete
0 ≤ FX (x) ≤ 1 ∀x ∈ SX , (11.13)
FX ( − ∞) = 0, (11.14)
FX (∞) = 1. (11.15)
Theorem 11.3. The cdf of a random variable X is an increasing function over the
whole range of x in the sample space, i.e.,
FX (x1 ) ≤ FX (x2 ) ∀x1 ≤ x2 ∈ SX . (11.16)
Proof. The proof follows from Definition 11.5 with x1 ≤ x2 , i.e.,
FX (x1 ) = Pr [X ≤ x1 ] ≤ Pr [X ≤ x2 ] = FX (x2 )].

Theorem 11.4. The probability that a discrete random variable X is within a specified
interval can be determined via its cdf and pmf as follows:
Pr [a < X ≤ b] = FX (b) − FX (a), (11.17)
Pr [a ≤ X ≤ b] = FX (b) − FX (a) + fX (a), (11.18)
Pr [a < X < b] = FX (b) − FX (a) − fX (b), (11.19)
Pr [a ≤ X < b] = FX (b) − FX (a) + fX (a) − fX (b). (11.20)
Proof. The proof follows from Definitions 11.5 and 11.3 as
Pr [a < X ≤ b] = Pr [X ≤ b] − Pr [X ≤ a]
= FX (b) − FX (a),
Pr [a ≤ X ≤ b] = Pr [X ≤ b] − Pr [X ≤ a] + Pr [X = a]
= FX (b) − FX (a) + fX (a),
Pr [a < X < b] = Pr [X ≤ b] − Pr [X ≤ a] − Pr [X = b]
= FX (b) − FX (a) − fX (b),
Pr [a ≤ X < b] = Pr [X ≤ b] − Pr [X ≤ a] + Pr [X = a] − Pr [X = b]
= FX (b) − FX (a) + fX (a) − fX (b).
11.2.3 Expected value

The expected value, a.k.a. mean, of a discrete random variable can be simply under-
stood as its average value over a number of tests or trials. Statistically, it is defined as
follows:
Definition 11.6. The expected value of a discrete random variable X is a statistical
probability-weighted average value of all possible outcomes, i.e.,

E[X ] xfX (x), (11.21)
x∈SX
where E[·] denotes the expectation operator, and fX (x) is the pmf of X given by (11.4).
In statistics, the expected value can also be represented by μX . For consistency,
the expected value is denoted by E[X ] throughout.
Example 11.4. Consider Example 11.1 with the pmf of random variable X given by
⎧
⎪
⎨1/4 if x = 0, 2,
fX (x) = 1/2 if x = 1,
⎪
⎩
0 otherwise.
The expected value of X is calculated by using (11.21) as

E[X ] = xfX (x) = 0 × 1/4 + 1 × 1/2 + 2 × 1/4 = 1.
x∈{0,1,2}
Theorem 11.5. The expected value of a derived discrete random variable Y = g(X )
is given by

E[Y ] = g(x)fX (x), (11.22)
x∈SX
where fX (x) is the pmf of X given by (11.4).
Proof. The proof follows from Definition 11.6 and Theorem 11.2 as

E[Y ] = yfY (y)
y∈SY

= g(x) fX (g −1 (y))
y∈SY x:x=g −1 (y)

= g(x)fX (x)
y∈SY x:x=g −1 (y)

= g(x)fX (x).
x∈SX
Lemma 11.1. The expected value of a random variable Y = aX + b is given by

E[Y ] = E[aX + b] = aE[X ] + b. (11.23)
Proof. From Theorem 11.5, we have

E[Y ] = g(x)fX (x)
x∈SX

= (ax + b)fX (x)
x∈SX

=a xfX (x) + b fX (x)
x∈SX x∈SX
= aE[X ] + b,

where the last equal sign is due to (11.6), i.e., x∈SX fX (x) = 1, and
Definition 11.6.
Considering multiple random variables, we have the following theorems:
Theorem 11.6. The expected value of a linear combination of N arbitrary random

variables {X1 , X2 , . . . , XN } is given by
N
N
E[Y ] = E ai Xi = ai E[Xi ]. (11.24)
i=1 i=1

N
E[Y ] = ai fXi (xi )
i=1 xi ∈SXi

N
= ai E[Xi ].
i=1
Theorem 11.7. If two random variables X and Y are independent, then

E[XY ] = E[X ]E[Y ]. (11.25)
Proof. From Definition 11.6, we have

E[XY ] = xy Pr [X = x, Y = y]
x,y∈SX ,Y

= x Pr [X = x] y Pr [Y = y]
x∈SX y∈SY
= E[X ]E[Y ].
where the second equal sign is based on Definition 10.7 of two independent
events.
11.2.4 Variance and standard deviation

Besides expected value, it is vital to determine variance of a discrete random variable.
The variance of a random variable is used to show how far the outcomes are stretched
from the average value. Statistically, it is defined as follows:
Definition 11.7. The variance of a discrete random variable X is a statistical average
value of the squared distance between the outcomes and their expected value, i.e.,

V [X ] E (X − E[X ])2 (11.26)

= (x − E[X ])2 fX (x), (11.27)
x∈SX
where V [X ] denotes the variance of X , E[X ] is the expected value of X given by

(11.21), and fX (x) is the pmf of X given by (11.4).
In statistics, the expected value can also be represented by σX2 . For consistency,
the variance value is denoted by V [X ] throughout.
Example 11.5. Consider Example 11.1 with the pmf of random variable X given by
⎧
⎪
⎨1/4 if x = 0, 2,
fX (x) = 1/2 if x = 1,
⎪
⎩
0 otherwise.
and the expected value of X is E[X ] = 1. The variance of X is calculated by using
(11.27) as

V [X ] = (x − E[X ])2 fX (x)
x∈{0,1,2}
= (0 − 1)2 1/4 + (1 − 1)2 1/2 + (2 − 1)2 1/4 = 1/2.
Theorem 11.8. The variance of a random variable X can be determined by

V [X ] = E[X 2 ] − E 2 [X ], (11.28)
where E[X ] is the expected value of X given by (11.21) and E[X 2 ] is given by

E[X 2 ] = x2 fX (x). (11.29)
x∈SX
Proof. From (11.26) in Definition 11.7, we have

V [X ] = E (X − E[X ])2

= E X 2 − 2XE[X ] + E 2 [X ]
(a)
= E[X 2 ] − 2E[X ]E[X ] + E 2 [X ]
= E[X 2 ] − E 2 [X ],
where (a) follows from Lemma 11.1.
Note that E[X 2 ] is different from E 2 [X ].
Theorem 11.9. The variance of a derived discrete random variable Y = g(X ) is

given by

V [Y ] = (g(x) − E[g(X )])2 fX (x), (11.30)
x∈SX
Lemma 11.2. The variance of a random variable Y = aX + b is given by

V [Y ] = V [aX + b] = a2 V [X ]. (11.31)

V [Y ] = (g(x) − E[g(X )])2 fX (x)
x∈SX
(a)
= (ax + b − (aE[X ] + b))2 fX (x)
x∈SX

= a2 (x − E[X ])2 fX (x)
x∈SX
(b)
= a2 V [X ],
where (a) follows from Lemma 11.1 and (b) follows from Definition 11.7.
The variance of a random variable can also be represented via a standard deviation
which is simply defined as the square root of its variance.
Definition 11.8. The standard deviation of a random variable X is the square root
of its variance, i.e.,

σX V [X ], (11.32)
where σX denotes the standard deviation of X , V [X ] is the variance of X given by
either (11.26) or (11.28).
√ of random variable X is V [X ] = 1/2,

Example 11.6. In Example 11.5, the variance
so the standard deviation of X is σX = 1/ 2.
Considering multiple random variables, there is a new term called covariance to
reflect the relationship between two random variables. The covariance is defined as
follows:
Definition 11.9. The covariance of two random variables X and Y , denoted by

cov[X , Y ], is defined as
cov[X , Y ] E[(X − E[X ])(Y − E[Y ])]. (11.33)
Theorem 11.10. The covariance of two random variables X and Y can be

determined by
cov[X , Y ] = E[XY ] − E[X ]E[Y ]. (11.34)
cov[X , Y ] = E[(X − E[X ])(Y − E[Y ])]
= E[XY − XE[Y ] − YE[X ] + E[X ]E[Y ]]
= E[XY ] − E[X ]E[Y ].
Lemma 11.3. If X and Y are independent, then cov[X , Y ] = 0
Proof. The proof follows from Theorems 11.7 and 11.10.
Theorem 11.11. The variance of a linear combination of N arbitrary random

variables {X1 , X2 , . . . , XN } is given by
N
N
N
V [Y ] = V a i Xi = a2i V [Xi ] + 2 ai aj cov[Xi , Xj ]. (11.35)
i=1 i=1 i=1 j
=i∈[1,N ]
N

V [Y ] = V ai X i
i=1
⎡ N 2 ⎤
N
= E⎣ ai X i − ai E[Xi ] ⎦
i=1 i=1
N

=E a2i (Xi − E[Xi ]) 2
i=1

N
+2 ai aj E[(Xi − E[Xi ])(Xj − E[Xj ])]
i=1 j
=i∈[1,N ]

N
N
= a2i V [Xi ] +2 ai aj cov[Xi , Xj ].
i=1 i=1 j
=i∈[1,N ]
Lemma 11.4. If random variables {X1 , X2 , . . . , XN } are independent, then

N
N
V [Y ] = V a i Xi = a2i V [Xi ]. (11.36)
i=1 i=1
Proof. The proof follows from Theorem 11.11, and the fact in Lemma 11.3 that
cov[Xi , Xj ] = 0 if Xi and Xj are independent.
11.2.5 Moments
In order to model a generalized change of a probability distribution of a random
variable about its average value, a new term, namely, central moment, is widely used
in statistics. The central moment and its variant forms of a discrete random variable
are defined as follows:
Definition 11.10. The nth central moment of a discrete random variable X is

E [(X − E[X ])n ] = (x − E[X ])n fX (x), (11.37)
x∈SX

It can be noticed in (11.37) that the second central moment, i.e., n = 2, is the
variance of X .
Definition 11.11. The nth moment, a.k.a. raw moment, of a discrete random variable
X is

E[X n ] = xn fX (x), (11.38)
x∈SX

The raw moment is actually the nth central moment about zero. It can be noticed
in (11.38) that the first moment, i.e., n = 1, is the expected value of X .
The moment is also represented in a normalized form, namely, normalized
moment, which is defined as follows:
Definition 11.12. The nth normalized moment, a.k.a. standardized moment, of a
discrete random variable X is

E [(X − E[X ])n ] x∈SX (x − E[X ]) fX (x)
n
= n , (11.39)
σX
n
(x − E[X ]) 2 f (x)
x∈SX X
where σX is the standard deviation of X given by (11.32) and fX (x) is the pmf of X
given by (11.4).
Specifically, the normalized moment has been applied in analyzing the shape of
a probability distribution. For instance,
● When n = 3, the third normalized moment is called skewness which is a measure
of the asymmetry of the probability distribution of a random variable about its
expected value.
● When n = 4, the fourth normalized moment is called kurtosis which is a measure
of the heaviness of the tail of the probability distribution of a random variable.
11.3 Continuous random variables

Definition 11.13. A random variable X : S → R is a continuous random variable if
the range of X is an uncountable set of numerical values in a specified interval either
closed, half-bounded, or open, i.e.,
SX = (x1 , x2 ) or (x1 , x2 ] or [x1 , x2 ) or [x1 , x2 ]. (11.40)
Note that the probability of any individual outcome of a continuous random

variable is zero, i.e., Pr [X = x] = 0, ∀x ∈ SX . This means that there is no pmf for
continuous random variables. Instead, a continuous random variable is represented
by a probability density function (pdf) and a cdf.
11.3.1 Cumulative distribution function

The cdf of a continuous random variable is defined as follows:
Definition 11.14. The cdf FX (x) of a continuous random variable X is
FX (x) Pr [X ≤ x]. (11.41)
0.9
0.8
0.7
0.6
FX (x)
0.5
0.4
0.3
0.2
0.1
0
0 0.5 1 1.5 2 2.5 3 3.5 4
x
Figure 11.4 cdf of a continuous uniform random variable in [1, 3]
Example 11.7. cdf of a continuous uniform random variable in the range [1, 3] as
shown in Figure 11.4 is given by
⎧
⎪
⎪ 0 if x ≤ 1,
⎨
x−1
FX (x) = if 1 ≤ x ≤ 3,
⎪
⎪
⎩ 2
1 if x > 3.
Remark 11.3. Note that the cdf given by (11.41) is identical to that for discrete
random variable as in (11.11); however, it is not computed by (11.12) which is only
feasible for a finite set. Due to the continuity of the real values in the sample space,
the cdf of a continuous random variable is also a continuous function.
Remark 11.4. It can be shown that the cdf of a continuous random variable also has
the properties of the discrete random variables, i.e.,
0 ≤ FX (x) ≤ 1 ∀x ∈ SX , (11.42)
FX ( − ∞) = 0, (11.43)
FX (∞) = 1, (11.44)
FX (x1 ) ≤ FX (x2 ) ∀x1 ≤ x2 ∈ SX . (11.45)
11.3.2 Probability density function

In order to describe the density of a continuous random variable, the pdf is defined
as follows:
Definition 11.15. The pdf fX (x) of a continuous random variable X is
dFX (x)
fX (x) = , (11.46)
dx
where FX (x) is the cdf of X determined by (11.41).
1.5
1
fX (x)
0.5
–0.5
1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3
x
Figure 11.5 pdf of a continuous uniform random variable in [1, 3]
Example 11.8. pdf of a continuous uniform random variable in the range [1, 3] as
shown in Figure 11.5 is determined by
⎧
⎨1
if 1 ≤ x ≤ 3,
fX (x) = 2
⎩0 otherwise.
Remark 11.5. From Definition 11.15, it can be noticed that the pdf of a continuous
fX (x) ≥ 0 ∀x ∈ SX , (11.47)
∞
fX (x)dx = 1, (11.48)
−∞
Theorem 11.12. The cdf of a continuous random variable X can be determined via
its pdf as
x
FX (x) = fX (t)dt, (11.49)
−∞
where fX (t) is the pdf of X .
Proof. The proof follows from Definitions 11.5 and 11.15, i.e.,
x
FX (x) = Pr [X ≤ x] = fX (t)dt.
−∞
Theorem 11.13. The probability that a continuous random variable X is within a

specified interval can be determined via its cdf and pdf as follows:
Pr [a < X ≤ b] = FX (b) − FX (a) (11.50)
b
= fX (x)dx. (11.51)
a
Proof. The proof of (11.50) is similar to that of Theorem 11.4, and the proof of
(11.51) follows from (11.49) in Theorem 11.12, i.e.,
Pr [a < X ≤ b] = FX (b) − FX (a)

b a
= fX (x)dx − fX (x)dx
−∞ −∞
b
= fX (x)dx.
a
Remark 11.6. Note that the interval between a and b in Theorem 11.13 can be either
left-bounded, right-bounded, closed, or open since Pr [X = x] = 0, ∀x ∈ SX , i.e.,
Pr [a ≤ X < b] = Pr [a ≤ X ≤ b] = Pr [a < X < b] = FX (b) − FX (a). (11.52)
11.3.3 Expected value

The expected value, a.k.a. mean, of a continuous random variable is defined as
follows:
Definition 11.16. The expected value of a continuous random variable X is
determined by
∞
E[X ] xfX (x)dx, (11.53)
−∞
where fX (x) is the pdf of X given by (11.46).
Theorem 11.14. The expected value of a derived continuous random variable Y =

g(X ) is given by
∞
E[Y ] g(x)fX (x)dx, (11.54)
−∞
11.3.4 Variance and standard deviation

The variance and standard deviation of a continuous random variable are defined as
follows:
Definition 11.17. The variance of a continuous random variable X is a statisti-
cal average value of the squared distance between the outcomes and their expected
value, i.e.,

V [X ] E (X − E[X ])2 (11.55)
∞
= (x − E[X ])2 fX (x)dx, (11.56)
−∞
where E[X ] is the expected value of X given by (11.53) and fX (x) is the pdf of X given
by (11.46).
Theorem 11.15. The variance of a continuous random variable X can be deter-

mined by
V [X ] = E[X 2 ] − E 2 [X ], (11.57)
where E[X ] is the expected value of X given by (11.53) and E[X 2 ] is given by
∞
E[X ] =
2
x2 fX (x)dx. (11.58)
−∞

Note that E[X 2 ] is different from E 2 [X ].
Theorem 11.16. The variance of a derived continuous random variable Y = g(X ) is

given by
∞
V [Y ] = (g(x) − E[g(X )])2 fX (x)dx, (11.59)
−∞

Similarly, the standard deviation of a continuous random variable is defined as
the square root of its variance.
Definition 11.18. The standard deviation of a continuous random variable X is the
square root of its variance, i.e.,

σX V [X ], (11.60)
where σX denotes the standard deviation of X and V [X ] is the variance of X given
by either (11.56) or (11.57).
11.3.5 Moments
The central moment and its variant forms of a continuous random variable are defined
as follows:
Definition 11.19. The nth central moment of a continuous random variable X is
∞
E [(X − E[X ]) ] = n
(x − E[X ])n fX (x)dx, (11.61)
−∞

Definition 11.20. The nth moment, a.k.a. raw moment, of a continuous random
variable X is
∞
E[X ] =
n
xn fX (x)dx, (11.62)
−∞

Definition 11.21. The nth normalized moment, a.k.a. standardized moment, of a

continuous random variable X is
∞
E [(X − E[X ])n ] −∞
(x − E[X ])n fX (x)dx
= ∞ n , (11.63)
σXn (x − E[X ]) 2 f (x)dx
−∞ X
where σX is the standard deviation of X and fX (x) is the pdf of X given by (11.46).
Remark 11.7. Similar to the discrete random variable, the first raw moment, the
second central moment, the third normalized moment, and the fourth normalized
moment of the continuous random variable stand for its variance, expected value,
skewness, and kurtosis, respectively.
11.4 Related works

references: Feller in 1968 [235], Prohorov and Rozanov in 1969 [243], Papoulis in
2002 [215], Yates and Goodman in 2004 [236], Johnson et al. in 1994 and 2005
[244,245], Gubner in 2006 [237], Leon-Garcia in 2007 [238], Stewart in 2009 [246],
[247], Forbes et al. in 2010 [216], and Mitzenmacher and Upfal in 2017 [242].
[Q11.1] What is a random variable in an experiment?

[Q11.2] Provide two examples to differentiate between discrete and continuous
random variables.
[Q11.3] What are the probability mass function (pmf) and cumulative distribution
function (cdf) of a discrete random variable?
[Q11.4] What are the properties of pmf and cdf of a discrete random variable?
[Q11.5] Show that the cdf of a random variable X is an increasing function over
the whole range of values in the sample space.
[Q11.6] Derive the formula to compute the probability that a discrete random
variable is within a specified interval.
[Q11.7] Derive the pmf of a derived discrete random variable Y = g(X ).
[Q11.8] What is the meaning of expected value, variance, and standard deviation
of a random variable?
[Q11.9] How to determine the expected value and variance of a discrete random
variable given its pmf?
[Q11.10] Derive the expected value of a linear sum of N random variables.
[Q11.11] Derive the expected value of a product of N independent random variables.
[Q11.12] Prove that the variance of a random variable X can be computed through
the first and the second moments of X , i.e., V [X ] = E[X 2 ] − E 2 [X ].
[Q11.13] Derive the expected value and variance of a derived random variable Y =
g(X ).
[Q11.14] What is covariance of two random variables and its meaning? Draw the
condition of two independent random variables.
[Q11.15] Derive the variance of a linear sum of N independent random variables.
[Q11.16] What are the nth central moment, the nth moment, and the nth normalized
moment of a discrete random variable?
[Q11.17] What is the probability density function (pdf) of a continuous random
variable and its relation with the cdf?
[Q11.18] Differentiate between the pdf and the pmf. Explain why the continuous
random variable cannot be represented by the pmf.
[Q11.19] Derive the probability that a continuous random variable is within a
specified interval via its pdf and cdf.
[Q11.20] How to determine the expected value and variance of a discrete random
variable given its pdf?
[Q11.21] Demonstrate the differences between the continuous and discrete random
variables in calculating the expected value, variance, and moments.
11.6 Problems

[P11.1] A random variable Z is given by Z = Ni=1 ai Xi + M j=1 bj Yj + c. The
expected value and variance of Xi , i = 1, 2, . . . , N , and Yj , j = 1, 2, . . . , M ,
are E[Xi ] = μXi , V [Xi ] = σX2i , E[Yj ] = μYj , and V [Yj ] = σY2j .
● Derive the general expressions for the expected value and variance
of Z.
● Find E[Z] and V [Z] for the following special cases:
– cov[Xi , Xj ] = 0, ∀i
= j, {i, j} ∈ {1, 2, . . . , N }.
– cov[Yi , Yj ] = 0, ∀i
= j, {i, j} ∈ {1, 2, . . . , M }.
– cov[Xi , Yj ] = 0, ∀i ∈ {1, 2, . . . , N } and j ∈ {1, 2, . . . , M }.
[P11.2] Consider the data transmission between two devices A and B. Assume that
A transmits three packets, each of which can be either successful (s) or
failed (f ) to receive at B with the same probability.
● Find the sample space of this experiment.
● Let X denote the number of packets that are successfully received
at B. Derive the probability mass function (pmf) and the cumulative
distribution function (cdf) of X .
● Find the probability that there is at least one packet successfully
received.
● Find the probability that there is either one or two packets successfully
received.
● Find the expected value, variance, and standard deviation of X .

● Find the skewness and kurtosis of X .
[P11.3] A discrete random variable X has the following pmf:

c/x if x = 1, 2, 3,
fX (x) =
0 otherwise,
where c is a constant.
● Find the value of constant c.
● Find Pr [X ≥ 2].
● Find the expected value and variance of X .
[P11.4] A discrete random variable X has the following pmf:
⎧
⎪
⎨1/8 if x = 0, 2,
fX (x) = 3/8 if x = 1, 3,
⎪
⎩
0 otherwise.
● Derive the cdf of X .

● Find Pr [0 ≤ X ≤ 2].
● Let Y = 3X + 1. Derive the pmf and cdf of Y .
● Find the expected value and variance of Y .
● Find the covariance of X and Y .
[P11.5] The probability that a call is a voice call is Pr [V ] = 0.7. The probability
of a data call is Pr [D] = 0.3. Voice calls cost 25 pence each and data calls
cost 40 pence each. Let C denote the cost (in pence) of one telephone call.
● Derive the pmf of C.
● Find the expected value and variance of C.
[P11.6] Monitor three phone calls and observe whether each one is a voice call or
a data call. The number of voice calls N has the following pmf
⎧
⎪
⎨0.1 if n = 0,
fN (n) = 0.3 if n = 1, 2, 3,
⎪
⎩
0 otherwise.
Voice calls cost 25 pence each and data calls cost 40 pence each. Let
T (pence) denote the cost of the three telephone calls monitored in the
experiment.
● Find the expected value and variance of N .
● Express T as a function of N .
● Derive the pmf of T.
● Find the expected value and variance of T based on those of N .
[P11.7] In an experiment to monitor two calls, the pmf of the number of voice
calls, i.e., N , is
⎧
⎪
⎪ 0.1 if n = 0,
⎪
⎨ 0.4 if n = 1,
fN (n) =
⎪
⎪ 0.5 if n = 2,
⎪
⎩
0 otherwise.
● Find the expected value and variance of N .
● Find the skewness and kurtosis of N .
[P11.8] The number of memory chips M needed in a personal computer depends on
how many application programs, A, the owner wants to run simultaneously.
The number of chips M and the number of application programs A are
described by
⎧
⎪
⎨4 if A = 1, 2,
M = 6 if A = 3,
⎪
⎩
8 if A = 4.

0.1(5 − a) if a = 1, 2, 3, 4,
fA (a) =
0 otherwise.
● Find the expected number of programs.

● Express M as a function of A.
● Find the expected number of memory chips.
[P11.9] The cdf of a continuous random variable X is given by
⎧
⎪
⎨0 if x < 0,
FX (x) = cx if 0 ≤ x ≤ 4,
⎪
⎩
1 if x > 4.
● Find the constant c.
● Find Pr [X ≤ 1].
● Find Pr [2 < X ≤ 3].
● Find Pr [X > 3/2].
● Derive the probability density function (pdf) of X .
[P11.10] A continuous random variable X has the following pdf

cxe−x/2 if x ≥ 0,
fX (x) =
0 otherwise.
● Find the constant c.
● Find Pr [0 ≤ X ≤ 4].
● Find Pr [ − 2 ≤ X ≤ 2].
[P11.11] The pdf of a random variable X is

3x2 /2 if − 1 ≤ x ≤ 1,
fX (x) =
0 otherwise.
● Find Pr [0 ≤ X ≤ 1] using the pdf.
● Derive the cdf of X . Verify Pr [0 ≤ X ≤ 1] with the derived cdf.
[P11.12] A triangular distributed random variable X has the following pdf:
⎧
⎪
⎨x − 1 if 1 ≤ x ≤ 2,
fX (x) = 3 − x if 2 < x 3,
⎪
⎩
0 otherwise.
Chapter 12
Discrete random variables and their applications
Discrete random variables, as introduced in Chapter 11, have been well developed
to model a number of practical network phenomena which are vital for network
simulation. This chapter will discuss various families of discrete random variables
and their applications in modeling and validating random events.
In this chapter, I will first provide a review of formulas related to discrete ran-
dom variables in Section 12.1. Then, various kinds of discrete random variables,
including Bernoulli, binomial, geometric, Pascal, Poisson, and discrete uniform ran-
dom variables, will be sequentially presented in Sections 12.2–12.7. Specifically, the
probability mass function (pmf), expected value, variance, and properties of each of
these random variables will be derived and discussed in detail.
12.1 Review of discrete random variables

A discrete random variable X : S → R has a countably infinite sample space
SX = {x1 , x2 , . . . }
or a finite sample space
SX = {x1 , x2 , . . . , xN }.
Its distribution functions, properties, and related formulas are summarized in
Table 12.1, where fX (x), FX (x), E[X ], V [X ], and σX denote the pmf, cumulative
distribution function, expected value, variance, and standard deviation, respectively,
of the random variable X .
12.2 Bernoulli random variable

Bernoulli random variable is named for a random variable with only two possible
outcomes. Let us consider the following example.
Example 12.1. Perform a test of a network device. There are only two possible
outcomes which could be either a success or a failure. This is also known as boolean
or binary outcomes. The test is successful with probability p and failed with probability
1 − p. Let a random variable X represent the number of successful devices in a test,
where X has the value of 1 if the test is successful and 0 if the test is failed.
Table 12.1 Formulas of discrete random variables
pmf fX (x) Pr [X = x]
0 ≤ fX (x) ≤ 1

x∈S X fX (x) = 1
fX |B (x) Pr [X = x|B]

fX (x) = M i=1 fX |B (x) Pr [Bi ]

cdf FX (x) Pr [X ≤ x] = x ≤x∈S X fX (x )
0 ≤ FX (x) ≤ 1, FX (−∞) = 0, FX (∞) = 1
FX (x1 ) ≤ FX (x2 ) ∀x1 ≤ x2
Pr [a < X ≤ b] = FX (b) − FX (a) ∀a < b

E[X ] E[X ] x∈S X xfX (x)

V [X ] V [X ] E (X − E[X ])2 = x∈S X (x − E[X ])2 fX (x)
V [X ] = E[X 2 ] − E 2 [X ]

σX σX V [X ]

Y = g(X ) fY (y) = x:x=g −1 (y) fX (g −1 (y))

E[Y ] = x∈S X g(x) fX (x)
E[aX + b] = aE[X ] + b

V [Y ] = x∈S X (g(x) − E[g(X )])2 fX (x)
V [aX + b] = a2 V [X ]

Central moment E [(X − E[X ])n ] = x∈SX (x − E[X ])n fX (x)

Raw moment E[X n ] = x∈SX xn fX (x)

Normalized moment (E [(X − E[X ])n ])/σXn = x∈SX (x − E[X ]) fX (x) /
n

n
x∈SX (x − E[X ]) fX (x)
2
In Example 12.1, the random variable X is called Bernoulli random variable,

and the experiment is well known as Bernoulli trial. For simplicity, it is written as
X ∼ Ber(p), and its distribution function is defined as follows:
Definition 12.1. The pmf of a Bernoulli random variable X ∼ Ber(p), 0 < p < 1, is
⎧
⎪
⎨p if x = 1,
fX (x) = 1 − p if x = 0, (12.1)
⎪
⎩
0, otherwise.
Discrete random variables and their applications 199
An illustration of a Bernoulli random variable X ∼ Ber(0.6) is shown in

Figure 12.1.
0.6
0.5
0.4
fX (x)
0.3
0.2
0.1
0
0 1
x
Figure 12.1 pmf of X ∼ Ber(0.6)

Remark 12.1. It can be observed in (12.1) that x∈SX fX (x) = 1. This reflects the
fact that the total pmf should be equal to 1 (cf. (11.6)). Note that this fact applies to
all discrete random variables including those covered in this chapter.
Theorem 12.1. The expected value and variance of a Bernoulli random variable
X ∼ Ber(p) are respectively given by
E[X ] = p, (12.2)
V [X ] = p(1 − p). (12.3)
E[X ] = 1 × fX (1) + 0 × fX (0)

= p,
V [X ] = (1 − p)2 fX (1) + (0 − p)2 fX (0)
= (1 − p)2 p + p2 (1 − p)
= p(1 − p),
12.3 Binomial random variable

As a generalization of Bernoulli random variable, binomial random variable represents
either the number of successes or the number of failures in a sequence of independent
Bernoulli trials. Let us consider the following example.
Example 12.2. Perform the test of N network devices, each of which is successful
with probability p and failed with probability (1 − p). The test result of a device is
independent of those of other devices. Let a random variable X represent the number
of successful devices in the N tests.
In Example 12.2, each individual test is a Bernoulli trial as in Example 12.1,

and the random variable X is named binomial random variable. For simplicity, it is
written as X ∼ Bin(N , p).
Definition 12.2. The pmf of a binomial random variable X ∼ Bin(N , p), 0 < p < 1,
N ∈ Z+ , is
⎧
⎪
⎨
N x
p (1 − p)N −x if x = 0, 1, 2, . . . , N ,
fX (x) = x (12.4)
⎪
⎩
0 otherwise,

N
where denotes the binomial coefficient given by
x

N N!
, (12.5)
x x!(N − x)!
and k! denotes the factorial of a nonnegative integer k computed by

k
k! i. (12.6)
i=1
An illustration of a binomial random variable X ∼ Bin(20, 0.6) is shown in

Figure 12.2.

N
Remark 12.2. In (12.4), the binomial coefficient is indeed the number of ways
x
to select x items out of N items, and thus it is read as “N choose x.” In this case,
this is the number of successful items out of N tests, each of which has a successful
probability of p.
Remark 12.3. A binomial random variable is a Bernoulli random variable when

N = 1.
0.18
0.16
0.14
0.12
0.1
fX (x)
0.08
0.06
0.04
0.02
0
0 2 4 6 8 10 12 14 16 18 20
x
Figure 12.2 pmf of X ∼ Bin(20, 0.6)

N N
Remark 12.4. Note that 0! = 1, and thus = = 1, which means fX (x)
0 N
still exists when x = 0 and x = N . In these two special cases, fX (0) = (1 − p)N and
fX (N ) = pN .
Considering the total pmf, we have the following theorem:
Theorem 12.2. For 0 < p < 1 and N ∈ Z+ ,

N
N x
p (1 − p)N −x = 1. (12.7)
x=0
x
Proof. The proof follows from the well-known binomial formula for the expansion
of (x + y)n based on the Pascal’s triangle, i.e.,
n
n
(x + y)n = xn−k yk .
k=0
k
Applying to the pmf given by (12.4), we have
N
N x
p (1 − p)N −x = (1 − p + p)N = 1.
x=0
x
Theorem 12.3. The expected value and variance of a binomial random variable
X ∼ Bin(N , p) are respectively given by
E[X ] = Np, (12.8)
V [X ] = Np(1 − p). (12.9)
Proof. The proof follows from the fact that the binomial random variable can be
regarded as the summation of N Bernoulli random variables, i.e.,

N
X = Xi ,
i=1
where Xi , i = 1, 2, . . . , N , is a Bernoulli random variable.

From Theorems 11.11 and 12.1 with E[Xi ] = p, i = 1, 2, . . . , N , the expected
value of X can be obtained by

N
E[X ] = E[Xi ]
i=1
= Np.
Since {X1 , X2 , . . . , XN } are independent, from Lemma 11.4 with the variance of each
Xi , i = 1, 2, . . . , N , given by V [Xi ] = p(1 − p) (cf. Theorem 12.1), the variance of X
can be determined by

N
V [X ] = V [Xi ]
i=1
= Np(1 − p).
12.4 Geometric random variable

A geometric random variable is used to represent the number of tests required to
obtain the first success.1 Let us consider the following example.
Example 12.3. Perform a test of network devices, each of which is successful with
probability p and failed with probability (1 − p). Let X be the number of tests up to
and including the first successful test.
In Example 12.3, the random variable X is called geometric random variable
due to the fact that its function distribution has the form of a geometric progression
1
The geometric random variable may be defined differently in some materials as the number of tests until
having the first success or before the first success. It is also defined as the number of failures prior to
having the first success. In either case, the distribution function can be similarly obtained by shifting the
random variable.
as shown later in Definition 12.3. For simplicity, the geometric random variable is
written as X ∼ Geo(p).
Definition 12.3. The pmf of a geometric random variable X ∼ Geo(p), 0 < p < 1, is
⎧
⎨p(1 − p)x−1 if x = 1, 2, . . . ,
fX (x) = (12.10)
⎩0 otherwise.
An illustration of a geometric random variable X ∼ Geo(0.6) is shown in
Figure 12.3.
0.6
0.5
0.4
fX (x)
0.3
0.2
0.1
0
2 4 6 8 10 12 14 16 18 20
x
Figure 12.3 pmf of X ∼ Geo(0.6)
Remark 12.5. In (12.10), it can be easily seen that there is only one success with
probability p, while the remaining (x − 1) tests are failed with probability (1 − p).
Remark 12.6. If we define a random variable Y as the number of failures until having
the first success, i.e., Y = X − 1, then the pmf of Y should be
⎧
⎨p(1 − p) y if y = 0, 1, 2, . . . ,
fY (y) = (12.11)
⎩0 otherwise.
In this case, there is y failures with probability (1 − p) and one success with probability
p. Notice that the range of y is from 0 for the case of no failure, which is different from
the range of random variable X in Definition 12.3 starting from 1.
Considering the total pmf, we have the following theorem
Theorem 12.4. For 0 < p < 1,

∞
p(1 − p)x−1 = 1. (12.12)
x=1
Proof. We have

∞
∞
p(1 − p)x−1 = p (1 − p)x−1
x=1 x=1

∞
= p (1 − p)x
x=0
(a) 1
= p
1 − (1 − p)
= 1,
where (a) follows from the infinite sum of geometric series, i.e.,

∞
1
qx = , ∀|q| < 1.
x=0
1−q
Theorem 12.5. The expected value and variance of a geometric random variable
X ∼ Geo(p) are respectively given by
1
E[X ] = , (12.13)
p
1−p
V [X ] = . (12.14)
p2
Proof. From Definition 11.6, the expected value of X is given by

E[X ] = xfX (x)
x∈SX
(a)
∞
= xp(1 − p)x−1
x=1

∞
= p x(1 − p)x−1
x=1
(b) 1
= p
(1 − (1 − p))2
1
= .
p
where (a) follows from the pmf of X in (12.3) and (b) follows from the derivative of
geometric progression, i.e.,

∞
1
nxn−1 = , ∀|x| < 1.
x=1
(1 − x)2
In order to derive the variance of X , we need to first compute E[X 2 ] as

E[X 2 ] = x2 fX (x)
x∈SX

∞
= x2 p(1 − p)x−1
x=1

∞
∞
= p x(x + 1)(1 − p)x−1 − p x(1 − p)x−1
x=1 x=1
(a) 2 1
= p −
(1 − (1 − p))3 p
2 1
= 2− ,
p p
where (a) follows from the derivative of geometric progression, i.e.,

∞
2
n(n + 1)xn−1 = , ∀|x| < 1.
x=1
(1 − x)3
From Theorem 11.8, we can obtain the variance of X as
V [X ] = E[X 2 ] − E 2 [X ]
2 1 1
= 2− − 2
p p p
1−p
= .
p2
12.5 Pascal random variable

A Pascal random variable is generalized from geometric random variable to represent
the number of tests required to obtain a specified number of successes. Let us consider
the following example.
Example 12.4. Perform a test of network devices, each of which is successful with
probability p and failed with probability (1 − p). Let X be the number of tests up to
and including the k successful tests.
In Example 12.4, the random variable X is called Pascal random variable and,
for simplicity, it is written as X ∼ Pas(k, p).
Definition 12.4. The pmf of a Pascal random variable X ∼ Pas(k, p), 0 < p < 1,
k ∈ Z+ , is
⎧
⎪ x−1 k
⎨ p (1 − p)x−k if x = k, k + 1, . . . ,
fX (x) = k − 1 (12.15)
⎪
⎩
0 otherwise.
An illustration of a Pascal random variable X ∼ Pas(4, 0.6) is shown in
Figure 12.4.
0.25
0.2
0.15
fX (x)
0.1
0.05
0
4 6 8 10 12 14 16 18 20
x
Figure 12.4 pmf of X ∼ Pas(4, 0.6)

x−1
Remark 12.7. In (12.15), the binomial coefficient is the number of possi-
k −1
bilities to have (k − 1) successful tests out of (x − 1) tests given the fact that the last
test is successful, i.e., the kth successful test.
Remark 12.8. It can be seen that the binomial coefficient in the pmf of Pascal random
variable has a reversed form of that of the binomial random variable where the number
of tests is fixed. Therefore, the Pascal random variable is also called as “negative”
binomial random variable in some materials.
Remark 12.9. A Pascal random variable is a geometric random variable when k = 1.

Considering the total pmf, we have the following theorem
Theorem 12.6. For 0 < p < 1 and N ∈ Z+ ,

∞
x−1 k
p (1 − p)x−k = 1. (12.16)
x=k
k − 1
Proof.
∞
∞

x−1 x−1
pk (1 − p)x−k = pk (1 − p)x−k
x=k
k −1 x=k
k −1
∞
1 x−1
= pk (1 − p)x−1
(1 − p)k−1 x=k k − 1
(a) 1 (1 − p)k−1
= pk
(1 − p)k−1 (1 − (1 − p))k
= 1,
where (a) follows from the ordinary generating function of binomial coefficients, i.e.,
∞
n xm
xn =
n=m
m (1 − x)m+1
when n = x − 1, m = k − 1, and x = 1 − p.
Theorem 12.7. The expected value and variance of a Pascal random variable X ∼
Pas(k, p) are respectively given by
k
E[X ] = , (12.17)
p
k(1 − p)
V [X ] = . (12.18)
p2
Proof. The proof follows from the fact that the Pascal random variable can be regarded
as the summation of k geometric random variables, i.e.,

k
X = Xi ,
i=1
where Xi , i = 1, 2, . . . , k, is a geometric random variable.

From Theorems 11.11 and 12.5 with E[Xi ] = 1/p, i = 1, 2, . . . , N , the expected

k
E[X ] = E[Xi ]
i=1
k
= .
p
Since {X1 , X2 , . . . , Xk } are independent, from Lemma 11.4 with the variance of each
Xi , i = 1, 2, . . . , N , given by V [Xi ] = (1 − p)/p2 (cf. Theorem 12.5), the variance of
X can be determined by

k
V [X ] = V [Xi ]
i=1
k(1 − p)
= .
p2
12.6 Poisson random variable
A Poisson random variable is associated with Poisson point process which was
described in Chapter 9 (cf. Section 9.2). In this section, the Poisson random variable
is discussed in detail showing its pmf, expected value, and variance.
A Poisson random variable X is used to represent the number of events occurring
in a specified time interval. For simplicity, X is written as X ∼ Poi(λ) where λ is a
Poisson parameter. As shown later, λ is also the average number of events, and thus
it is also known as event rate.
Definition 12.5. The pmf of a Poisson random variable X ∼ Poi(λ), λ > 0 is

⎧ x −λ
⎨λ e if x = 0, 1, 2, . . . ,
fX (x) = x! (12.19)
⎩
0 otherwise.
An illustration of a Poisson random variable X ∼ Poi(5) is shown in Figure 12.5.

Considering the total pmf, we have the following theorem:
Theorem 12.8. For λ > 0,

∞
λx e−λ
= 1. (12.20)
x=0
x!
0.18
0.16
0.14
0.12
0.1
fX (x)
0.08
0.06
0.04
0.02
0
0 2 4 6 8 10 12 14 16 18 20
x
Figure 12.5 pmf of X ∼ Poi(5)
Proof. We have

∞
λx e−λ
∞
λx
= e−λ
x=0
x! x=0
x!
(a)
= e−λ eλ
= 1,
where (a) follows from the Maclaurin series of exponential function, i.e.,

∞
xn
ex = .
n=0
n!
Theorem 12.9. The expected value and variance of a Poisson random variable X ∼
Poi(λ) are respectively given by
E[X ] = λ, (12.21)
V [X ] = λ. (12.22)

E[X ] = xfX (x)
x∈SX
(a) ∞
λx e−λ
= x
x=0
x!
∞
λx e−λ
= x
x=1
x!
∞
λx−1
= λe−λ
x=1
(x − 1)!

∞
λx
= λe−λ
x=0
x!
(b)
= λe−λ eλ
= λ
where (a) follows from the pmf of X given by (12.19) and (b) follows from the
Maclaurin series of exponential function, i.e.,

∞
xn
ex = .
n=0
n!
From Theorem 11.8, the variance of X is determined by

V [X ] = E[X 2 ] − E 2 [X ]
∞
λx e−λ
= x2 − λ2
x=0
x!

∞
λx e−λ
= x − λ2
x=0
(x − 1)!

∞
λx−1
= λe−λ x − λ2
x=0
(x − 1)!

∞
λx−1
= λe−λ x − λ2
x=1
(x − 1)!

∞
λ x−1 ∞
λ x−1
= λe−λ (x − 1) + − λ2
x=1
(x − 1)! x=1
(x − 1)!
∞
λx−2 ∞
λx−1
−λ
= λe λ + − λ2
x=2
(x − 2)! x=1
(x − 1)!
(a)
= λe−λ (λeλ + eλ ) − λ2
= λ,
where (a) follows the same Maclaurin series of exponential function as in the proof
of the expected value of X .
12.7 Discrete uniform random variable

A discrete uniform random variable is used to describe an experiment with equiprob-
able outcomes within a finite range of integer numbers.
Let X denote the discrete uniform random variable with the sample space SX =
{a, a + 1, . . . , b}, where a, b ∈ Z and a < b. The random variable X can be simply
written as X ∼ U (a, b).
Definition 12.6. The pmf of a discrete uniform random variable X ∼ U (a, b), a, b ∈
Z, a < b, is
⎧
⎪ 1
⎨ if x = a, a + 1, a + 2, . . . , b,
fX (x) = b−a+1 (12.23)
⎪
⎩
0 otherwise.
An illustration of a discrete uniform random variable X ∼ U (1, 5) is shown in

Figure 12.6.
0.2
0.18
0.16
0.14
0.12
fX (x)
0.1
0.08
0.06
0.04
0.02
0
1 2 3 4 5
x
Figure 12.6 pmf of X ∼ U (1, 5)

Remark 12.10. It can be observed in (12.23) that

b
1
= 1.
x=a
b−a+1
This verifies that the total pmf should be equal to 1 (cf. (11.6)).
Theorem 12.10. The expected value and variance of a discrete uniform random
variable X ∼ U (a, b) is
a+b
E[X ] = , (12.24)
2
(b − a + 1)2 − 1
V [X ] = . (12.25)
12

E[X ] = xfX (x)
x∈SX
(a)
b
1
= x
x=a
b−a+1
1 b
= x
b − a + 1 x=a
(b) 1 (b − a + 1)(a + b)
=
b−a+1 2
a+b
= ,
2
where (a) follows from the pmf of X given by (12.23) and (b) follows from the sum
of series that

b
(b − a + 1)(a + b)
i= .
i=a
2
In order to derive the variance of X , for simplicity, let us define another random
variable Y = X − a + 1. Y is also a uniform random variable, but it is defined within
a simple range, i.e., Y ∼ U (1, n), where n = b − a + 1. The pmf of Y is thus given by
⎧
⎪
⎨
1
if y = 1, 2, . . . , n,
fY (y) = n
⎪
⎩0 otherwise.
The expected value of Y can be obtained as

E[Y ] = yfY (y)
y∈SY
n
1
= y
y=1
n
1
n
= y
n y=1
n+1
= .
2
Since Y is a linear combination of X , we have another approach to derive E[X ] using

Theorem 11.6 as
E[X ] = E[Y ] + a − 1
b−a+2
= +a−1
2
a+b
= .
2
In order to derive V [Y ], we need to first compute E[Y 2 ] as follows

E[Y 2 ] = y2 fY (x)
y∈SY

n
1
= y2
y=1
n
1 2
n
= y
n y=1
1 n(n + 1)(2n + 1)
(a)
=
n 6
(n + 1)(2n + 1)
= ,
6
where (a) follows from the sum of sequence of squares, i.e.,

n
n(n + 1)(2n + 1)
i2 = .
i=1
6
The variance of Y can be derived from Theorem 11.8 as
V [Y ] = E[Y 2 ] − E 2 [Y ]
(n + 1)(2n + 1) (n + 1)2
= −
6 4
n −1
2
=
12
(b − a + 1)2 − 1
= .
12
According to Lemma 11.2, we have
(b − a + 1)2 − 1
V [X ] = V [Y ] = .
12
12.8 Related works
references:
Bernoulli Distribution Johnson et al. in 2005 [245] and Forbes et al. in 2010 [216].
Binomial Distribution Miller in 1954 [248], Kaas and Buhrman in 1980 [249], and
Wang in 1993 [250].
Geometric Distribution Andrews in 1998 [251] and Spiegel in 2003 [252].
Pascal Distribution Wilf in 2006 [253] and Hilbe in 2011 [254].
Poisson Distribution Haight in 1967 [255], Knusel in 1986 [256], Daley and Vere-
Jones in 2003 and 2008 [221,222], Illian et al. in 2008 [223], Bass in 2011 [224],
and Snyder and Miller in 2012 [225].
Discrete Uniform Distribution Papoulis in 2002 [215], Johnson et al. in 2005 [245],
and Forbes et al. in 2010 [216].
[Q12.1] Give an example for the following discrete random variables:

● Bernoulli random variable.
● Binomial random variable.
● Geometric random variable.
● Pascal random variable.
● Poisson random variable.
● Discrete uniform random variable.
[Q12.2] For each of the above random variables, describe its probability mass
function (pmf).
[Q12.3] Derive the expected value and variance of the above random variables.
[Q12.4] What is the relationship between binomial random variable and Bernoulli
random variable?
[Q12.5] What is the relationship between Pascal random variable and geometric
random variable?
[Q12.6] What is the physical meaning of the Poisson parameter in defining the
distribution function of a Poisson random variable?
12.10 Problems
[P12.1] A Bernoulli test of network devices is represented by a Bernoulli random
variable X ∼ Ber(0.4). Find its expected value and variance of X .
[P12.2] Perform a test of five devices, each of which follows a Bernoulli distribu-
tion with a success probability of 0.3. Assume the tests are independent
of each other.
● Derive the probability mass function (pmf) of the number of successful
devices denoted by X .
[P12.3] A test of devices has a successful probability of 0.6. Let X denote the
number of tests up to and including the last successful attempt.
● How to represent X ?
● Derive the probability mass function (pmf) of X .
[P12.4] Consider a test of network devices. Let X denote the number of failures
until the first success.
● Derive the pmf of X .
● What is the relationship between X and the geometric random
variable?
[P12.5] Each time a modem transmits 1 b, the receiving modem analyzes the signal
that arrives and decides whether the transmitted bit is 0 or 1. It makes an
error with probability p, independent of whether any other bit is received
correctly.
● If the transmission continues until the receiving modem makes its first
error, what is the pmf of X , the number of bits transmitted?
● If p = 0.1, what is the probability that X = 10? What is the probability
that X ≥ 10?
● If the modem transmits 100 b, what is the pmf of Y , the number of
errors?
● If p = 0.01 and the modem transmits 100 b, what is the probability
of Y = 2 errors at the receiver? What is the probability that Y ≤ 2?
● If the transmission continues until the receiving modem makes three
errors, what is the pmf of Z, the number of bits transmitted?
● If p = 0.25, what is the probability of Z = 12 b transmitted?
[P12.6] A device is rejected with probability of p = 1/4. Let X represent the

number of tests up to and including the first reject.
● Derive the pmf of X .
● Derive the cumulative distribution function (cdf) of X .
● Use the derived cdf to find the following probabilities:
– Pr [X < 1]
– Pr [X ≤ 1]
– Pr [X > 2]
– Pr [X ≥ 2]
– Pr [3 < X ≤ 8]
– Pr [X = 1]
– Pr [X = 3]
[P12.7] The number of hits at a website in any time interval is a Poisson random
variable. A particular site has on average λ = 2 hits per second.
● Derive the pmf of the Poisson random variable X ∼ Poi(λ).
● Find the probability that there are no more than two hits in an interval
of 1 s.
● Let Y denote the number of hits in 0.25 s. Derive the pmf of Y and
find the probability that there is no hit within that interval.
● Find the probability that there are at least three hits in a 2-s interval.
[P12.8] Consider information flow arriving at a router. Suppose that an average
of 50 flows per minute arrives at the router following a Poisson process.
Let X denote the number of arrivals during the next 5 s.
● Derive the pmf and cdf of X .
● Find the probability that three or more flows will arrive at the router
in the next 5 s.
[P12.9] The amplitude V [V] of a sinusoidal signal follows a discrete uniform
distribution within a range of value {−3, −2, −1, 0, 1, 2, 3}.
● Derive the pmf of V .
● Find the expected value and variance of V .
● Let W = V 2 /2 [W] denote the average power of the transmitted signal.
– Derive the pmf of W .
– Find the expected value and variance of W .
Chapter 13
Continuous random variables
and their applications
Together with discrete random variables for modeling discrete random events, contin-
uous random variables have also been well developed to represent network phenomena
in which the variables can randomly take infinite and uncountable values. This chapter
will discuss various families of continuous random variables and their applications
in modeling and validating random events.
In this chapter, I will first provide a review of formulas related to continuous
random variables in Section 13.1. Then, various kinds of continuous random variables,
including uniform, exponential, Erlang, normal, and lognormal random variables,
will be sequentially discussed in Sections 13.2–13.6. For each random variable, its
probability density function (pdf), cumulative distribution function (cdf), expected
value, and variance will be derived along with discussion of their properties and
relevant applications.
13.1 Review of continuous random variables

A continuous random variable X : S → R has an uncountable sample space consisting
of numerical values in a specified interval which can be either closed, half-bounded,
or open, i.e.,
SX = (x1 , x2 ) or (x1 , x2 ] or [x1 , x2 ) or [x1 , x2 ].
Its distribution functions, properties, and related formulas are summarized in
Table 13.1, where fX (x), FX (x), E[X ],V [X ], and σX denote the pdf, cdf, expected
value, variance, and standard deviation, respectively, of the random variable X .
13.2 Continuous uniform random variable
Similar to discrete uniform random variable in Chapter 12 (cf. Section 12.7), a con-
tinuous uniform random variable is used to represent an event with equiprobable
outcomes having values within a specified range.
Table 13.1 Formulas of continuous random variables
pdf fX (x) = ((dFX (x))/dx)

0 ≤ fX (x) ≤ 1
∞
−∞ fX (x) = 1 x
cdf FX (x) Pr [X ≤ x] = −∞ fX (t)dt
0 ≤ FX (x) ≤ 1, FX ( − ∞) = 0, FX (∞) = 1
FX (x1 ) ≤ FX (x2 ) ∀x1 ≤ x2
b
Pr [a < X ≤ b] = FX (b) − FX (a) = a fX (x)dx ∀a < b
∞
E[X ] E[X ] −∞ xfX (x)dx
∞
V [X ] V [X ] E (X − E[X ])2 = −∞ (x − E[X ])2 fX (x)dx
V [X ] = E[X ] − E [X ]
2 2
σX σX V [X ]
Y = g(X ) fY (y) = x:x=g −1 (y) fX (g −1 (y)
∞
E[Y ] = −∞ g(x) fX (x)dx
E[aX + b] = aE[X ] + b
∞
V [Y ] = −∞ (g(x) − E[g(X )])2 fX (x)dx
V [aX + b] = a2 V [X ]
∞
Central moment E [(X − E[X ])n ] = −∞ (x − E[X ])n fX (x)dx
∞
Raw moment E[X n ] = −∞ xn fX (x)dx
− E[X ]) ] )/σX )
n n
Normalized moment ((E [(X
∞
n
∞
= ( −∞ (x − E[X ])n fX (x)dx)/ −∞ (x − E[X ])2 f (x)dx
X
Let X denote the continuous uniform random variable with the sample space
SX = [a, b], where a < b. The random variable X can be simply written as X ∼
U (a, b) with a notice that the range of values is continuous.
Definition 13.1. The pdf of a continuous uniform random variable X ∼ U (a, b),
a < b, is
⎧
⎨ 1 if a ≤ x ≤ b,
fX (x) = b − a (13.1)
⎩0 otherwise
Remark 13.1. It can be shown from (13.1) that
∞ b
1
fX (x)dx = dx
b−a
−∞ a
1
= x|b
b−a a
= 1.
Continuous random variables and their applications 219
1.5
0.5
fX (x)
–0.5
–1
1 1.5 2 2.5 3 3.5 4 4.5 5
x
Figure 13.1 pdf of X ∼ U(1,5)
This verifies the fact that the total pdf should be equal to 1 (cf. (11.48)). Note that
this fact applies to all continuous random variables including those covered in this
chapter.
An illustration of a continuous uniform random variable is shown in Figures 13.1

and 13.2 where the pdf and cdf of X ∼ U (1, 5) are, respectively, plotted.
Theorem 13.1. The cdf of a continuous uniform random variable X ∼ U (a, b), a < b,
is given by
⎧
⎪
⎪ 0 if x < a
⎨x − a
FX (x) = if a ≤ x ≤ b (13.2)
⎪
⎪ b−a
⎩
1 if x > b
Proof. The proof follows from Definition 11.14 and Theorem 11.12 as follows:
● If x ≤ a, then
FX (x) = Pr [X ≤ x] = 0.
0.9
0.8
0.7
0.6
FX (x)
0.5
0.4
0.3
0.2
0.1
0
0 1 2 3 4 5 6
x
Figure 13.2 cdf of X ∼ U(1,5)
● If a ≤ x ≤ b, then
x
FX (x) = fX (t)dt
−∞
x
1
= fX (t)dt
b−a
a
x

t
=
b − a a
x−a
= .
b−a
● If x > b, then
FX (x) = Pr [X ≤ x] = 1.
Theorem 13.2. The expected value and variance of a continuous uniform random
variable X ∼ U (a, b) are respectively given by
a+b
E[X ] = , (13.3)
2
(b − a)2
V [X ] = . (13.4)
12
∞
E[X ] = xfX (x)dx
−∞
b
(a) x
= dx
b−a
a
b
1 x2
=
b−a 2 a
a+b
= ,
2
where (a) follows from (13.1).

From Theorem 11.15, in order to derive the variance of X , we need to calculate
E[X 2 ] which is given by
∞
E[X ] =
2
x2 fX (x)dx
−∞
b
x2
= dx
b−a
a
b
1 x3
=
b−a 3 a
a2 + ab + b2
= .
3
The variance is thus obtained by
V [X ] = E[X 2 ] − E 2 [X ]
2
a2 + ab + b2 a+b
= −
3 2
(b − a)2
= .
12
13.3 Exponential random variable

An exponential random variable is related to Poisson point process (PPP) which was
briefly discussed in Chapter 9 (cf. Section 9.2). As stated in Theorem 9.1, the time
difference between events, a.k.a. interarrival time, is exponentially distributed.
The exponential random variable X is represented by a parameter λ > 0, namely,
rate as in PPP. For simplicity, it is generally written as X ∼ Exp(λ).
Definition 13.2. The pdf of an exponential random variable X ∼ Exp(λ), λ > 0, is

λe−λx if x ≥ 0
fX (x) = (13.5)
0 otherwise.
Considering the total pdf, we have the following theorem
Theorem 13.3. For λ > 0,

∞
λe−λx = 1. (13.6)
0
Proof. We have
∞ ∞
1
λe−λx = λ − e−λx
λ 0
0
(a) 1
= λ
λ
= 1,
where (a) follows from the fact that
lim e−λx = 0.
x→∞
An illustration of an exponential random variable is shown in Figures 13.3

and 13.4 where the pdf and cdf of X ∼ Exp(5) are, respectively, plotted.
Theorem 13.4. The cdf of an exponential random variable X ∼ Exp(λ), λ > 0, is

given by

1 − e−λx if x ≥ 0
FX (x) = (13.7)
0 otherwise.
Proof. The proof follows from Definition 11.14 and Theorem 11.12 as follows:
● If x < 0, then
FX (x) = Pr [X ≤ x] = 0.
4.5
3.5
3
fX(x)
2.5
1.5
0.5
0
0 0.5 1 1.5 2 2.5 3
x
Figure 13.3 pdf of X ∼ Exp(5)
0.9
0.8
0.7
0.6
0.5
FX (x)
0.4
0.3
0.2
0.1
0
0 0.5 1 1.5 2 2.5 3
x
Figure 13.4 cdf of X ∼ Exp(5)

● If x ≥ 0, then
x
FX (x) = fX (t)dt
−∞
x
= λe−λt dt
0
x
= −e−λt 0
= 1 − e−λx .
Theorem 13.5. The expected value and variance of an exponential random variable
X ∼ Exp(λ) are respectively given by
1
E[X ] = , (13.8)
λ
1
V [X ] = . (13.9)
λ2
∞
E[X ] = xfX (x)dx
−∞
∞
(a)
= xλe−λx dx
0
∞
∞
= = −xe−λx 0 +
(b)
e−λx dx
0

(c) 1 −λx ∞
= − e
λ 0
1
= ,
λ
where (a) follows from (13.5), (b) follows by performing integration by parts, and (c)
follows from the fact that
lim xe−λx = 0.
x→∞
From Theorem 11.15, in order to derive the variance of X , we need to calculate

E[X 2 ] which is given by
∞
E[X ] =
2
x2 fX (x)dx
−∞
∞
= x2 λe−λx dx
0
∞
∞
= −x2 e−λx 0 +
(a)
2xe−λx dx
0
(b) 2
= E[X ]
λ
(c) 2
= 2,
λ
where (a) follows by performing integration by parts, and (b) follows from the proof
of E[X ] with a fact that
lim x2 e−λx = 0
x→∞
and (c) follows from (13.8).

The variance is thus obtained by
V [X ] = E[X 2 ] − E 2 [X ]
2
2 1
= 2−
λ λ
1
= .
λ2
Theorem 13.6. An exponential random variable has a memoryless property, i.e.,
Pr [X ≤ a + b|X > a] = Pr [X ≤ b] (13.10)

Proof. From Definition 10.6 of conditional probability, we have
Pr [X ≤ a + b, X > a]
Pr [X ≤ a + b|X > a] =
Pr [X > a]
Pr [a < X ≤ a + b]
=
Pr [X > a]
(a) F X (a + b) − FX (a)
=
f − FX (a)
(b) 1 − e−λ(a+b) − (1 − e−λa )
=
1 − (1 − e−λa )
= 1 − e−λb
(c)
= FX (b)
(d)
= Pr [X ≤ b],
where (a) follows from Theorems 11.13 and 10.3, (b) and (c) follow from
Theorem 13.4, and (d) follows from Definition 11.14.
13.4 Erlang random variable

An Erlang random variable has the distribution of a sum of independent exponential
variables having the same expected values. It is used to represent the total interarrival
time in various random processes such as in queuing systems with PPP (cf. Section 9.2
in Chapter 9).
The Erlang random variable X is represented by two parameters including rate,
denoted by λ > 0, as for exponential random variable and shape, denoted by N ,
which corresponds the number of exponential random variables. For simplicity, it is
generally written as X ∼ Erl(N , λ).
Definition 13.3. The pdf of an Erlang random variable X ∼ Erl(N , λ), N ∈ Z+ ,
λ > 0, is
⎧ N N −1 −λx
⎨λ x e if x ≥ 0,
fX (x) = (N − 1)! (13.11)
⎩
0 otherwise.
Theorem 13.7. For N ∈ Z+ and λ > 0,

∞
λN xN −1 e−λx
dx = 1 (13.12)
(N − 1)!
0
Proof. We have
∞ ∞
λN xN −1 e−λx λN
dx = xN −1 e−λx dx
(N − 1)! (N − 1)!
0 0
λ (N − 1)!
N
=
(N − 1)! λN
= 1,
where the second equality follows from
∞
m!
xm e−ax dx =
am+1
0
when m = N − 1 and a = λ.
An illustration of an Erlang random variable is shown in Figures 13.5 and 13.6

where the pdf and cdf of X ∼ Erl(3, 5) are, respectively, plotted.
1.4
1.2
0.8
fX(x)
0.6
0.4
0.2
0
0 0.5 1 1.5 2 2.5 3
x
Figure 13.5 pdf of X ∼ Erl(3,5)

0.9
0.8
0.7
0.6
FX (x)
0.5
0.4
0.3
0.2
0.1
0
0 0.5 1 1.5 2 2.5 3
x
Figure 13.6 cdf of X ∼ Erl(3,5)

Theorem 13.8. The cdf of an Erlang random variable X ∼ Erl(N , λ), N ∈ Z+ , λ > 0,
is given by

N −1
(λx)k
FX (x) = 1 − e−λx (13.13)
k=0
k!
Proof. Let X1 , X2 , . . . , XN , denote N exponential random variables,

each having rate
λ > 0. The Erlang random variable X is thus given by X = Ni=1 Xi .
Related to PPP, the Erlang distribution also means the waiting time, i.e., λx,
until the N th Poisson event occurs. Let Y denote the Poisson random variable, i.e.,
Y ∼ Poi(λx). Therefore, in order to have a waiting time longer than x, all N − 1
Poisson random events should occur, i.e.,

N −1
Pr [X > x] = Pr [Y = k]
k=0
(a)
N −1
= fY (k)
k=0
(b)
N −1
(λx)k e−λx
=
k=0
k!

N −1
(λx)k
−λx
= e ,
k=0
k!
where (a) follows from Definition 11.3 with fY (k) denoting the probability mass
function (pmf) of Y and (b) follows from Definition 12.5 of the pmf of Poisson
random variable.
From Definition 11.14, the cdf of X is given by
FX (x) = Pr [X ≤ x]
= 1 − Pr [X > x]

N −1
(λx)k
= 1 − e−λx .
k=0
k!
Theorem 13.9. The expected value and variance of an Erlang random variable X ∼
Erl(N , λ) are respectively given by
N
E[X ] = , (13.14)
λ
N
V [X ] = 2 . (13.15)
λ
Proof. The proof follows from the fact that the Erlang random variable can be regarded
as the summation of N exponential random variables, i.e.,

N
X = Xi ,
i=1
where Xi , i = 1, 2, . . . , N , is an exponential random variable.

From Theorems 11.11 and 13.5 with E[Xi ] = 1/λ, i = 1, 2, . . . , N , the expected

N
E[X ] = E[Xi ]
i=1
N
= .
λ
Since {X1 , X2 , . . . , XN } are independent, from Lemma 11.4 with the variance of each
Xi , i = 1, 2, . . . , N , given by V [Xi ] = 1/λ2 (cf. Theorem 13.5), the variance of X can
be determined by

N
V [X ] = V [Xi ]
i=1
N
= .
λ2
13.5 Normal random variable

A normal random variable, a.k.a. Gaussian random variable, is a well-known family
of continuous random variables to represent the unknown distributions in general. An
example of Gaussian random variable can be found in modeling the random envi-
ronmental noises in wireless communications, and thus they are also called Gaussian
noises. The development of the normal random variable is due to the central limit the-
orem which states that the summation of various independent distributions converges
to the normal distribution having a bell-shaped pdf defined as in Definition 13.4.
The normal random variable X is represented by two parameters including μ and
σ 2 . As shown later in Theorem 13.12, μ and σ 2 are the expected value and variance,
respectively, of X . For simplicity, the normal random variable X is generally written
as X ∼ N (μ, σ 2 ).
Definition 13.4. The pdf of a normal random variable X ∼ N (μ, σ 2 ) is

1 (x − μ)2
fX (x) = √ exp − . (13.16)
2πσ 2σ 2
Theorem 13.10. For all μ and σ ,

∞
1 (x − μ)2
√ exp − dx = 1. (13.17)
2π σ 2σ 2
−∞
Proof. We have
∞ ∞ 2
1 (x − μ)2 (a) 1 t
√ exp − dx = √ exp − dt
2π σ 2σ 2
2π 2
−∞ −∞
∞ 2
(b)2 t
= √ exp − dt
2π 2
0
⎛∞ ⎞1/2
2 ∞ 2
2⎝ u v
dv⎠
(c)
= exp − du exp −
π 2 2
0 0
⎛∞ ∞ ⎞1/2
2
2⎝ u +v 2
= exp − dudv⎠
π 2
0 0
⎛∞ ∞ ⎞1/2
2
2⎝ u + (wu)2
ududw⎠
(d)
= exp −
π 2
0 0
⎛∞ ⎞1/2
2 u→∞
2⎝ 1 u (1 + w2 )
= − exp − dw⎠
π 1 + w2 2 u=0
0
⎛∞ ⎞1/2

2⎝ 1
= dw⎠
π 1 + w2
0

2 1/2
= arctan (w)|∞
π 0

2 π
=
π 2
= 1,
where (a) follows by letting t = (x − μ)/σ , (b) follows from the fact that e−t /2 is an
2
even function, (c) follows by duplicating the same integration along with changing
variable t = u and t = v, and (d) follows by changing variable v = wu.
Remark 13.2. Denoting Y as the normal random variable with μ = 0 and σ = 1,

we have the simplest form of the pdf given by
2
1 y
fY (y) = √ exp − . (13.18)
2π 2
The random variable Y ∼ N (0, 1) is known as standard normal random variable, and
the general normal random variable X ∼ N (μ, σ 2 ) can be written by
X = μ + σY. (13.19)
An illustration of a Gaussian random variable is shown in Figures 13.7 and 13.8

where the pdf and cdf of X ∼ N (1, 9) are, respectively, plotted.
Theorem 13.11. The cdf of a normal random variable X ∼ N (μ, σ 2 ) is given by

x−μ
FX (x) = (13.20)
σ

1 x−μ
= 1 + erf √ , (13.21)
2 σ 2
where (x) is the cdf of the standard normal distribution, i.e.,
x 2
1 t 1 x
(x) √ exp − dt = 1 + erf √ (13.22)
2π 2 2 2
−∞
0.14
0.12
0.1
0.08
fX(x)
0.06
0.04
0.02
0
–10 –8 –6 –4 –2 0 2 4 6 8 10
x
Figure 13.7 pdf of X ∼ N(1,9)
0.9
0.8
0.7
0.6
FX (x)
0.5
0.4
0.3
0.2
0.1
0
–10 –8 –6 –4 –2 0 2 4 6 8 10
x
Figure 13.8 cdf of X ∼ N(1,9)

and erf(x) is the error function defined by

x
2
e−t dt.
2
erf(x) √ (13.23)
π
0
Proof. The proof follows from Theorem 11.12 as follows:

x
FX (x) = fX (t)dt
−∞
x

(a) 1 (t − μ)2 x−μ
= √ exp − dt =
2πσ 2σ 2 σ
−∞
√

((x−μ)/(σ 2))
(b) 1 1
= +√ e−u du
2 π
0

(c) 1 x−μ
= 1 + erf √ ,
2 σ 2
where (a) follows from the pdf of X given by (13.16) and √ the cdf (x) defined
by
0 (13.22), (b) follows by changing variable u = (t − μ)/(σ 2) and the fact that
−∞
fX (t)dt = 1/2, and (c) follows from the definition of error function in (13.23).
Theorem 13.12. The expected value and variance of a normal random variable
X ∼ N (μ, σ 2 ) are respectively given by
E[X ] = μ, (13.24)
V [X ] = σ . 2
(13.25)
Proof. Let us first find the expected value and variance of a standard normal random
variable Y ∼ N (0, 1) having the pdf given by (13.18).
From Definition 11.16, the expected value of Y is given by
∞
E[Y ] = yfY (y)dy
−∞
∞ 2
y y
= √ exp − dy
2π 2
−∞
= 0,
where the second equality is due to the fact that the integrand is an odd function.
From Theorem 11.15, in order to derive the variance of Y , we need to calculate

E[Y 2 ] which is given by
∞
E[Y ] =
2
y2 fY (y)dy
−∞
∞ 2
y2 y
= √ exp − dy
2π 2
−∞
∞ 2
(a)2 y
= √ y2 exp − dy
2π 2
0
√
(b) 2 π
= √
π (4/(2 2))
= 1,
where (a) follows from the fact that the integrand is an even function and (b) follows
from
∞ √
m −ax2 π
x e dx = (m+1)/2
4a
0
when m = 2 and a = 1/2.

The variance of Y is thus obtained by
V [Y ] = E[Y 2 ] − E 2 [Y ]
= 1.
As stated in Remark 13.2, the random variable X can be represented by X =

μ + σ Y . From Theorem 11.6 and Lemma 11.4, we have
E[X ] = μ + σ E[Y ]
= μ,
V [X ] = σ 2 V [Y ]
= σ 2.
13.6 Lognormal random variable

A lognormal random variable is a continuous random variable having natural logarithm
of its distribution function is a normal distribution. An example of lognormal random
variable can be found in modeling the change of received power caused by shadowing
in wireless communications.
The lognormal random variable X is also represented by two parameters includ-
ing μ and σ 2 . For simplicity, the lognormal random variable X is written as
X ∼ LN (μ, σ 2 ).
Definition 13.5. The pdf of a lognormal random variable X ∼ LN (μ, σ 2 ) is
⎧
⎨ √1 ( ln x − μ)2
exp − if x > 0,
fX (x) = x 2πσ 2σ 2 (13.26)
⎩
0 otherwise.
Theorem 13.13. For all μ and σ ,

∞
1 ( ln x − μ)2
√ exp − dx = 1. (13.27)
2πσ x 2σ 2
0
Proof. We have
∞ ∞ 2
1 ( ln x − μ)2 (a) 1 t
√ exp − 2
dx = √ exp − dt
2πσ x 2σ 2π 2
0 −∞

(b) 2 π
=
π 2
= 1,
where (a) follows by letting t = ( ln x − μ)/σ and (b) follows the same proof of
Theorem 13.10.
Remark 13.3. Let Y denote a normal random variable having mean μ and variance
σ 2 , i.e., Y ∼ N (μ, σ 2 ). By definition, X can be written as
X = eY . (13.28)
An illustration of a lognormal random variable is shown in Figures 13.9 and 13.10

where the pdf and cdf of X ∼ LN (1, 0.25) are, respectively, plotted.
0.5
0.45
0.4
0.35
0.3
fX (x)
0.25
0.2
0.15
0.1
0.05
0
0 1 2 3 4 5 6
x
Figure 13.9 pdf of X ∼ LN(1,0.25)
0.9
0.8
0.7
0.6
FX (x)
0.5
0.4
0.3
0.2
0.1
0
0 1 2 3 4 5 6
x
Figure 13.10 cdf of X ∼ LN(1,0.25)

Theorem 13.14. The cdf of a lognormal random variable X ∼ LN (μ, σ 2 ) is given by

ln x − μ
FX (x) = (13.29)
σ

1 ln x − μ
= 1 + erf √ , (13.30)
2 σ 2
where (x) and erf(x) are given by (13.22) and (13.23), respectively.
Theorem 13.15. The expected value and variance of a lognormal random variable
X ∼ LN (μ, σ 2 ) are respectively given by

σ2
E[X ] = exp μ + , (13.31)
2
V [X ] = e2μ+σ (eσ − 1).

2 2
(13.32)
Proof. As noted in Remark 13.3, X can be represented by X = eY where Y ∼

N (μ, σ 2 ). The expected value of X can be derived from the pdf of Y using
Theorem 11.14 as
∞
E[X ] = ey fy (y)dy
−∞
∞
(a) 1 (y − μ)2
= ey √ exp − dy
2πσ 2σ 2
−∞
∞
σ2 1 (y − μ)2 σ2
= exp μ + √ exp y − −μ− dy
2 2πσ 2σ 2 2
−∞
∞
σ 2
1 (y − μ − σ 2 )2
= exp μ + √ exp −
2 2πσ 2σ 2
−∞

(b) σ 2
= exp μ + .
2
where (a) follows from Definition 13.4 and (b) follows from the fact that the integrand
is the pdf of a normal random variable N (μ + σ 2 , σ 2 ).
From Theorem 11.15, we have

V [X ] = E[X 2 ] − E 2 [X ]
(a) 2
= E[e2Y ] − e2μ+σ
(b) 2 2
= e2μ+2σ − e2μ+σ
= e2μ+σ (eσ − 1),
2 2
where (a) follows from the proof of the expected value and (b) follows from the fact
that 2Y ∼ N (2μ, 4σ 2 ).
13.7 Related works

references:
Continuous Uniform Distribution Johnson et al. in 1994 [244], Papoulis in 2002
[215], Yates and Goodman in 2004 [236], and Forbes et al. in 2010 [216].
Exponential Distribution Feller in 1968 [235], Ross in 1995 [219], and Balakrish-
nan and Basu in 1996 [257].
Erlang Distribution Jambunathan in 1954 [258], Papoulis in 2002 [215], and Forbes
et al. in 2010 [216].
Normal Distribution Patel and Read in 1996 [259], Papoulis in 2002 [215], Spiegel
in 2003 [252], and Forbes et al. in 2010 [216].
Lognormal Distribution Aitchison et al. in 1957 [260], Crow and Shimizu in 1987
[261], and Forbes et al. in 2010 [216].

[Q13.1] Provide an example for the following continuous random variables:
● Continuous uniform random variable.
● Exponential random variable.
● Erlang random variable.
● Normal random variable.
● Lognormal random variable.
[Q13.2] Describe the probability density function, usage, and application of the
above random variables.
[Q13.3] Derive the cumulative distribution function, the expected value, and
variance for each of the above random variables.
[Q13.4] Show that the exponential random variable has memoryless property.
[Q13.5] What is the relationship between Erlang random variable and exponential
random variable?
[Q13.6] What is the relationship between lognormal random variable and normal
random variable?
13.9 Problems
[P13.1] A continuous random variable X has E[X ] = 3 and V [X ] = 9. Find the
probability density function (pdf) and cumulative distribution function (cdf)
of X when
● X is a continuous uniform random variable.
● X is an exponential random variable.
[P13.2] The probability that a telephone call lasts no more than t minutes is modeled
by the following cdf:

1 − e−t/3 if t ≥ 0,
FT (t) =
0 otherwise.
● Derive the pdf of the duration in minutes of a telephone conversation.
● Find the probability that a conversation will last between 2 and 4 min.
● Find the expected duration of a telephone call.
● Find the variance and standard deviation of the call duration.
● Find the probability that a call duration is within ±1 standard deviation
of the expected call duration.
[P13.3] An exponentially distributed random variable X has the following relation-
ship:
P[X ≤ 1] = 0.9P[X ≤ 2].
● Find the value of λ.
● Derive the pdf and cdf of X .
[P13.4] The lifetime X , in years, of a networking device is given by the following
pdf:

0.4e−0.4x if x ≥ 0,
fX (x) =
0 otherwise.
● Find the probability that this device is still alive after 5 years.
● Find the probability that the device dies between 3 and 6 years from
the time it is placed in the networking system.
[P13.5] A component has an exponential time-to-failure with a mean of 10,000 h.
● Derive the pdf and cdf of the time-to-failure of the component.
● The component has already been in operation for its mean life. Find
the probability that it will fail by 15,000 h.
● At 15,000 h, the component is still in operation. Find the probability
that it will operate for another 5,000 h.
[P13.6] Phone company A charges 0.15 per minute for telephone calls. For any
fraction of a minute at the end of a call, they charge for a full minute.
Phone Company B also charges 0.15 per minute. However, Phone Company
B calculates its charge based on the exact duration of a call. Find the
expected revenues per call for companies A and B when the duration of a
call in minutes is exponentially distributed with a rate λ = 1/3. (Note that,
if X is an exponential (λ) random variable, then K =
X is a geometric
(p) random variable with p = 1 − e−λ and E[K] = 1/p.)
[P13.7] The time intervals between hits on a web page from remote computers are
exponentially distributed, with a mean of 15 s.
● Derive the pdf and cdf of the duration of K hits.
● Find the probability that the third hit connection occurs after 30 s have
elapsed.
[P13.8] Consider the validation of a network component. There are four steps to
validate the quality of a component. Each step requires an average of 10
min. The validation time for each step is exponentially distributed, and the
steps are independent of each other.
● Let X denote the total validation time of all four steps. Derive the pdf
and cdf of X .
● Find the expected value and variance of the total validation time.
● Find the probability that the whole validation finishes in 30 min or less.
● Find the probability that the third step of the validation process occurs
after 15 min have elapsed.
[P13.9] A standard normal random variable X has expected value E[X ] = 0 and
variance V [X ] = 1. Let Y = 2 + 4X and Z = eY .
● Derive the pdf and cdf of Y .
● Derive the pdf and cdf of Z.
● Find the expected value and variance of Z.
Chapter 14
Random variable generation in network
simulation
In network modeling and simulation, as introduced in Chapter 11, the randomness of

phenomena is represented by various random variables which are specified by proba-
bility distribution functions. Specifically, various families of discrete and continuous
random variables were discussed in Chapters 12 and 13, respectively. The critical
issue in network modeling for performance evaluation is to build a random variable
generator to simulate a random event given its probability distribution function. This
chapter is devoted to illustrating some widely used techniques for generating random
variables.
In this chapter, I will first introduce inverse transform technique in Section 14.1.
Specifically, examples of applying the inverse transform technique to generate both
discrete and continuous random variables will be provided. Based on the developed
random generators, Section 14.2 will present transformation technique to generate
related random variables in the same family.
14.1 Inverse transform technique

Given distribution of random variables, inverse transform technique can be used to
generate random samples of various discrete and continuous random variables, such as
Bernoulli, geometric, discrete uniform, continuous uniform, and exponential random
variables.
In order to apply this technique, for discrete random variable, it is required
to know the probability mass function (pmf) and/or cumulative distribution func-
tion (cdf), while the generation of continuous random variables is developed based
on its probability density function (pdf) and/or cdf. Also, as shown later in Theo-
rem 14.1, this technique can only be applied to generate the random variables having
invertible cdf.
As a starting point for generating random variables of different families, it is
assumed that there exists a basic generator of a uniform random variable R in the
range [0, 1], i.e., R ∼ U (0, 1).
The inverse transform technique is motivated by noticing that the cdf FX (x)
of a random variable X is monotonically increasing function (cf. Theorem 11.3 in
Chapter 11) and 0 ≤ FX (x) ≤ 1 ∀x ∈ SX .
We have Theorem 14.1
Theorem 14.1. Given a uniform random variable R ∼ U (0, 1) and a generalized

inverse function FX−1 (r), 0 < r < 1, of FX (x) defined as
FX−1 (r) = inf {x : FX (x) ≥ r} , (14.1)
if a random variable X is determined by
X = FX−1 (R), (14.2)
then X has the cdf FX (x) .
Proof. We have
Pr [X ≤ x] = Pr [FX−1 (R) ≤ x]
= Pr [R ≤ FX (x)]
= FX (x),
where the last equality is due to the fact that Pr [R ≤ r] = r if R ∼ U (0, 1). This
according means that FX (x) is the cdf of X .
From Theorem 14.1, the inverse transform technique for generating samples of
a random variable X given its cdf FX (x) can be realized via the following steps:
● Step 1: Generate samples of a uniform random variable R ∼ U (0, 1).
● Step 2: Determine the generalized inverse function of cdf, i.e., FX−1 (r).
● Step 3: For every r, find x such that x = FX−1 (r)
Remark 14.1. Note that the generalized inverse function in (14.1) is used when the
cdf FX (x) is not continuous, and thus it can be employed to generate discrete random
variables via the following steps:
● Step 1: Generate samples of a uniform random variable R ∼ U (0, 1).
● Step 2: For every r, increase xi from i = 1 until FX (xi ) > r, then return x = xi .
Therefore, all discrete distributions can be generated via inverse-transform technique
as long as its pmf is specified.
In the following subsections, some examples of the inverse transform technique

will be provided for both discrete and continuous random variables that were studied
in Chapters 12 and 13.
14.1.1 Bernoulli random variable generation

Theorem 14.2. A Bernoulli random variable X ∼ Ber(p), 0 < p < 1, can be
generated by

0 if U ≤ p,
X = (14.3)
1 otherwise.
Random variable generation in network simulation 243
Proof. The proof follows from the pmf of X given by Definition 12.1, i.e.,
⎧
⎪
⎨p if x = 1,
fX (x) = 1 − p if x = 0, (14.4)
⎪
⎩
0, otherwise.
An illustration of generating Bernoulli samples X ∼ Ber(0.6) is shown in

Figure 14.1.
600
500
400
300
200
100
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Figure 14.1 Samples of X ∼ Ber(0.6)
14.1.2 Geometric random variable generation

Theorem 14.3. A geometric random variable X ∼ Geo(p), 0 < p < 1, can be
generated by

ln (R)
X = , (14.5)
ln (1 − p)
where R ∼ U (0, 1) and · is the ceiling function defined by
x = min{t ∈ Z|t ≥ x}. (14.6)
Proof. We first derive the cdf of X from its pmf, i.e., fX (x), defined in Definition
12.3 as

p(1 − p)x−1 if x = 1, 2, . . . ,
fX (x) = (14.7)
0 otherwise.
From Definition 11.5, the cdf of X is given by

FX (x) = Pr [X ≤ x]
= 1 − Pr [X > x]
= 1 − Pr [X ≥ x + 1]

∞
= 1− p(1 − p)k−1
k=x+1

∞
= 1 − p(1 − p)x (1 − p)k
k=0
(a) 1
= 1 − p(1 − p)x
1 − (1 − p)
= 1 − (1 − p)x ,
where (a) follows from the infinite sum of geometric series, i.e.,

∞
1
qx = , ∀|q| < 1.
x=0
1−q
Solving FX (X ) > 1 − R with a notice that X is discrete, the theorem is proved.

An illustration of generating geometric samples X ∼ Geo(0.6) is shown in
Figure 14.2.
600
500
400
300
200
100
0
1 2 3 4 5 6 7 8
Figure 14.2 Samples of X ∼ Geo(0.6)

14.1.3 Discrete uniform random variable generation

Theorem 14.4. A discrete uniform random variable X ∼ U (a, b), a, b ∈ Z, a < b,
can be generated by
X = a +
(b − a + 1)R, (14.8)
where R ∼ U (0, 1) and
· is the floor function defined by

x = max{t ∈ Z|t ≤ x}. (14.9)
Proof. Note that the pdf of X is given in Definition 12.6 as

⎧
⎨ 1
if x = a, a + 1, a + 2, . . . , b,
fX (x) = b − a + 1 (14.10)
⎩0 otherwise.
From Definition 11.5, the cdf of X can be derived as
⎧
⎪
⎪ 1 if x > b,
⎨x − a + 1
FX (x) = if x = a, a + 1, a + 2, . . . , b, (14.11)
⎪
⎪ b−a+1
⎩
0 if x < a.
Solving FX (X ) > R with a notice that X is discrete, the theorem is proved.
An illustration of generating discrete uniform samples X ∼ U (1, 5) and X ∈ Z
is shown in Figure 14.3.
250
200
150
100
50
0
1 1.5 2 2.5 3 3.5 4 4.5 5
Figure 14.3 Samples of X ∼ U(1,5), X ∈ Z

14.1.4 Continuous uniform random variable generation

Theorem 14.5. A continuous uniform random variable X ∼ U (a, b), a, b ∈ R, a < b,
can be generated by
X = a + (b − a)R, (14.12)
where R ∼ U (0, 1).
Proof. The proof follows from Theorem 14.1 where the cdf of X was derived in
Theorem 13.1 as
⎧
⎪
⎪ 0 if x ≤ a
⎨x − a
FX (x) = if a ≤ x ≤ b (14.13)
⎪
⎪ b−a
⎩
1 if x > b
Solving FX (X ) = R with 0 < R < 1 to find X in terms of R, we obtain
X = a + (b − a)R.
An illustration of generating continuous uniform samples X ∼ U (1, 5) is shown

in Figure 14.4.
120
100
80
60
40
20
0
1 1.5 2 2.5 3 3.5 4 4.5 5
Figure 14.4 Samples of X ∼ U(1,5)

14.1.5 Exponential random variable generation

Theorem 14.6. An exponential random variable X ∼ Exp(λ), λ > 0, can be
generated by
1
X = − ln (1 − R), (14.14)
λ
where R ∼ U (0, 1).
Proof. The proof follows from Theorem 14.1 where the cdf of X was derived in
Theorem 13.4 as

1 − e−λx if x ≥ 0
FX (x) = (14.15)
0 otherwise.
Solving FX (X ) = R with 0 < R < 1 to find X in terms of R, we obtain
1
X = − ln (1 − R).
λ
An illustration of generating exponential samples X ∼ Exp(5) is shown in

Figure 14.5.
600
500
400
300
200
100
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Figure 14.5 Samples of X ∼ Exp(5)

14.2 Transformation techniques

For some random variables, they can be simply generated based on their features and
properties instead of performing the inversion of the cdf. In this section, various trans-
formation techniques will be discussed through the generation of binomial, Pascal,
Erlang, and lognormal random variables.
14.2.1 Binomial random variable generation

From the definition of binomial random variable in Section 12.3, it is a generalized
form of Bernoulli random variable for a fixed number of tests, i.e., N . The approach to
generate the binomial random variable X ∼ Bin(N , p), 0 < p < 1, N ∈ Z+ is known
as composition technique which consists of the following steps:
● Step 1: Generate N independent uniform random variables R1 ∼ U (0, 1), R2 ∼

U (0, 1) ∼ U (0, 1), . . . , RN ∼ U (0, 1).
● Step 2: Find X which is the number of random variables Ri satisfying Ri < p.
An illustration of generating binomial samples X ∼ Bin(20, 0.6) is shown in

Figure 14.6.
300
250
200
150
100
50
0
4 6 8 10 12 14 16 18
Figure 14.6 Samples of X ∼ Bin(20,0.6)

14.2.2 Pascal random variable generation

In Section 12.5, a Pascal random variable X ∼ Pas(k, p), 0 < p < 1, k ∈ Z+ , is
extended from geometric random variable. Specifically, it is the summation of k
geometric random variables, i.e.,

k
X = Xi ,
i=1
where Xi ∼ Geo(p), i = 1, 2, . . . , k.
The approach to generate the Pascal random variable X ∼ Pas(k, p), 0 < p < 1,
k ∈ Z+ is known as convolution method, which can be carried out as follows:
● Step 1: Generate k independent geometric random variables X1 ∼ Geo(p), X2 ∼

Geo(p), . . . , Xk ∼ Geo(p) (cf. Section 14.1.2).
● Step 2: Determine X as

k
X = Xi .
i=1
An illustration of generating Pascal samples X ∼ Pas(4, 0.6) is shown in

Figure 14.7.
350
300
250
200
150
100
50
0
4 6 8 10 12 14 16
Figure 14.7 Samples of X ∼ Pas(4,0.6)

300
250
200
150
100
50
0
0 0.5 1 1.5 2 2.5
Figure 14.8 Samples of X ∼ Erl(3,5)
14.2.3 Erlang random variable generation

In Section 13.4, an Erlang random variable X ∼ Erl(N , λ), N ∈ Z+ , λ > 0, is
extended from exponential random variable. Specifically, it is the summation of N
exponential random variables, i.e.,

N
X = Xi ,
i=1
where Xi ∼ Exp(λ), i = 1, 2, . . . , k.
Employing the same approach to generate Pascal random variables, the Erlang
random variable X ∼ Erl(N , λ) can be carried out with convolution technique as
follows:
● Step 1: Generate N independent exponential random variables X1 ∼ Exp(λ), X2 ∼
Exp(λ), . . . , XN ∼ Exp(λ) (cf.
Section 14.1.5).
● Step 2: Determine X as X = Ni=1 Xi .
An illustration of generating Erlang samples X ∼ Erl(3, 5) is shown in
Figure 14.8.
14.2.4 Lognormal random variable generation

As noted in Section 13.6, a lognormal random variable X ∼ LN (μ, σ 2 ) can be
represented as
X = eY , (14.16)
where Y ∼ N (μ, σ 2 ), i.e., a normal random variable with mean μ and variance σ 2 .
250
200
150
100
50
0
–1 –0.5 0 0.5 1 1.5 2 2.5
Figure 14.9 Samples of Y ∼ N(1,0.25)
Another notice in Section 13.5 is that Y ∼ N (μ, σ 2 ) can be written by
Y = μ + σ Z, (14.17)
where Z ∼ N (0, 1) is a standard normal random variable.

The generation of the lognormal random variable X ∼ LN (μ, σ 2 ) can be realized
by employing direct transformation technique as follows:
● Step 1: Generate a standard normal random variable Z ∼ N (0, 1).

● Step 2: Determine Y as Y = μ + σ Z.
● Step 3: Determine X as X = eY .
An illustration of generating normal and lognormal samples Y ∼ N (1, 0.25) and

X ∼ LN (1, 0.25) is shown in Figures 14.9 and 14.10, respectively.
14.3 Related works

references: Box and Muller in 1958 [262], Devroye in 1981 and 1986 [263,264],
Barabesi and Pratelli in 2015 [265], Robert and Casella in 2005 [266], Rubinstein
and Kroese in 2007 [267], and Gentle in 2009 [247].
400
350
300
250
200
150
100
50
0
0 2 4 6 8 10 12
Figure 14.10 Samples of Y ∼ LN(1,0.25)

[Q14.1] Describe the basic concept of inverse transform technique to generate a
random variable for network simulation.
[Q14.2] A random variable X is represented by a cumulative distribution function
(cdf) FX (x). Explain steps to generate X for the following cases:
● X is a continuous random variable.
● X is a discrete random variable.
[Q14.3] Provide steps with proof to generate the following discrete random
variables:
● Discrete uniform random variable.
● Bernoulli random variable.
● Geometric random variable.
[Q14.4] Provide steps with proof to generate the following continuous random
variables:
● Continuous uniform random variable.
● Exponential random variable.
[Q14.5] Provide some examples of employing transformation technique to generate
the following random variables:
● Binomial random variable.
● Pascal random variable.
● Erlang random variable.
● Lognormal random variable.
14.5 Problems
[P14.1] Develop a generator for the following random variables:
● X ∼ Ber(0.5).
● X ∼ Geo(0.5).
● X ∼ U (2, 6), X ∈ Z (discrete).
● X ∼ U (2, 6) (continuous).
● X ∼ Exp(3).
● X ∼ Bin(10, 0.5).
● X ∼ Erl(4, 3).
[P14.2] Develop a random-variate generator for a random variable X having the
following cdf:

e2x /2 if x ≤ 0,
F(x) =
1 − e−2x if x > 0.
[P14.3] Develop a generation scheme for the triangular distribution with the
following pdf:
⎧
⎪ x−2
⎪
⎪ 2 if 2 ≤ x ≤ 3,
⎨
f (x) = 2 − x/3
⎪ if 3 < x ≤ 6,
⎪
⎪
⎩ 2
0 otherwise.
[P14.4] Develop a generator for the variable having the following cdf:
⎧
⎪
⎪ 0 if x ≤ −3,
⎪
⎪
⎪
⎪ 1 x
⎨ + if − 3 < x ≤ 0,
F(x) = 21 6x2
⎪
⎪
⎪
⎪ + if 0 < x ≤ 4,
⎪
⎪ 2 32
⎩1 if x > 4.
[P14.5] Develop a generator for a random variable X having the following cdf:
⎧
⎪
⎨0 if x < 0,
F(x) = x /16 if 0 ≤ x ≤ 2,
4
⎪
⎩
1 if x > 2.
[P14.6] Develop a generator for a random variable X having the following cdf:
⎧
⎪
⎪ 0 if x < 0,
⎪
⎨x/3 if 0 ≤ x ≤ 2,
F(x) =
⎪
⎪ 2/3 + (x − 2)/24 if 2 < x ≤ 10,
⎪
⎩
1 if x > 10.
[P14.7] Develop a generator for a Weibull random variable X ∼ Wei(λ, k) with

the following cdf:

1 − e−(x/λ)
k
if x ≥ 0,
F(x) =
0 if x < 0,
where λ > 0 and k > 0 denote the scale parameter and shape parameter,
respectively.
Chapter 15
Queuing theory for network modeling
and performance evaluation
As outlined in Chapter 9, statistical models are crucial for network simulation and
performance evaluation when most parameters and events in practical networks are
randomly distributed and variant over time following different distributions. A typical
application of the statistical models can be found in queuing systems to model the
waiting experiences. This statistical modeling of the waiting is well known as queuing
theory. This chapter is devoted to study the queuing theory in the context of computer
and communication networks.
In this chapter, I will first introduce the basic concept of queuing theory in
Section 15.1. The application of queuing theory in the form of queuing systems
for computer and communication networks will be presented in Section 15.2. As a
background of the queuing system, the Poisson point process will be then reviewed in
Section 15.3. Various performance measures and their relationship with Little’s law
will be discussed in Section 15.4. This chapter will be closed by introducing Kendall’s
notation for specifying queuing systems in Section 15.5.
15.1 An introduction to queuing theory

Queuing theory has been adopted to study the waiting lines in different areas, such as
telecommunications, computing, industrial engineering, data management, etc. The
queuing theory is shown to be eminently helpful in modeling and validating the per-
formance of a number of facilities and components within computer, communication,
manufacturing, and transport systems.
Applying queuing theory to computer and communication systems, a variety
of computer processors, disks, terminals, communication links, concentrators, and
memories can be modeled as queuing systems. The queuing theory is regarded as a
mathematical tool for describing the behavior of queuing or waiting lines in a system.
Presented in mathematical terms, the queuing theory allows us to validate various
performance measures of interest, such as waiting times, waiting lengths, response
times, busy periods, idle periods, blocking probability, etc.
The queuing theory is aimed to model and analyze a system verifying its ability
to provide services as demanded, for instance to minimize the response time and
delay with respect to various levels of demand. In order to achieve this, the behavior,
operation, and properties of all resources as well as relevant parameters within the
system need to be understood and considered when developing any queuing models.
Indeed, the queuing model is generally subject to constraints of available resources
along with contention between them.
As modeling of a waiting line, a queuing model simply represents a service
facility to be able to accommodate requests that are waiting to be served. Depending
on the number of servers, the queuing model can be categorized into two basic types
including the following:
● Single-server queue: In a single-server queue, packets are sequentially served one
at a time. If there are more than one packet arriving, then there is only one packet
to be served, while the remaining packets queue in a buffer. After the service of
a packet is complete, it leaves the system and thus the number of packets in the
system is reduced by one. The next packet in the queue will be served and the
number of packets in the queue is also reduced by one.
● Multi-server queue: In a multi-server queue, packets are served by multiple
servers in parallel. If the number of packet arrivals is less than the number of
servers, then some servers will be idle. Otherwise, the number of packets being
served is the number of servers, while the remaining packets will queue in a
buffer.
The above two queuing models will be discussed in detail in the subsequent Chapters
16 and 17.
In a practical communication network consisting of multiple nodes, these nodes
are generally interconnected via various communication links, and thus they experi-
ence multiple queues of different types, namely, a network of queues. In this case,
the network of queues can be regarded as an aggregation of various simple queuing
models.
15.2 Queuing systems in computer and communication

networks
In computer and communication networks, the contention of resources is a critical

issue due to their inability to support immediately all requests for different kinds of
services. This accordingly results in the development of queues, which cause sub-
stantial delays in obtaining the required services. Such delays, in some networks such
as telephone or mobile phone networks, may cause rejection or blocking of service
requests. Therefore, in queuing systems, there are three major constraints, including
reliability, capacity, and delay (cf. Chapter 5 for a brief introduction of these three
performance measures).
Basically, a queuing system consists of the following components (cf. Fig-
ure 15.1):
● Source: is the input of the system which can be either an infinite or a finite source.
Although a finite source with a finite queue is more appropriate in practice, the
Queuing theory for network modeling and performance evaluation 257
Data packets
Packet arrival
Queue Server
Source
Figure 15.1 Components in a queuing system
source is generally assumed to be very large that can be approximated as an

infinite source for simplicity in the analysis.
● Packet arrivals: The packets are assumed to arrive at the queuing system randomly
over time and may occur according to any arbitrary pattern. Poisson arrivals are
shown to be completely random in time, and thus Poisson process is applicable
for modeling queuing systems in many applications. The average interarrival time
of packets is the reciprocal of the arrival rate.
● Service mechanism: depends on the number of servers as categorized above for
two basic queues including single-server and multi-server queues. The service is
represented by a service rate, which is the reciprocal of the average service time.
● Queue capacity: is related to the buffer size which allows the packets to queue
in. We therefore have the following:
– Zero capacity: the packets drop when system is busy, i.e., no buffer, and thus
it is also known as a loss system.
– Infinite capacity: all packets are allowed to enter the queue with unlimited
buffer size.
– Positive capacity: only a limited number of packets are allowed to enter the
queue if there is still room.
● Queue discipline: defines the rule that the server selects the next packet from the
queue, for instance first-in first-out (FIFO) or first-come first-served (FCFS),
last-in first-out (LIFO), first-in random-out (FIRO), etc.
15.3 Poisson point process for queuing systems

In this section, Poisson point process (PPP) will be reviewed with a brief summary
of related formula that will be used in this chapter and the following Chapters 16 and
17. Details of the Poisson random variables and Poisson point process can be referred
to Section 9.2 in Chapter 9 and Section 12.6 in Chapter 12.
The PPP in queuing systems is defined in one dimension to represent a counting

process. The PPP is related to Markov process as a special case in which the future
probabilities are determined by the current state of the process, but not on how that
state was reached. The PPP is also known as a birth–death process where the packet
arrival at the queue is “birth” and the service at the server is “death.” As shown in
Section 9.2, the PPP has the same memoryless property as defined in the Markov
process.
Let X denote a Poisson random variable, i.e., X ∼ Poi(λ), to represent the number
of events occurring in a specified time interval with a Poisson parameter λ > 0. The
probability mass function (pmf) of a Poisson random variable X is given by Definition
12.5, i.e.,
⎧ x −λ
⎨λ e
if x = 0, 1, 2, . . . ,
fX (x) = x! (15.1)
⎩
0 otherwise.
It has been shown in Theorem 12.9 that the expected value and variance of X are
E[X ] = V [X ] = λ. (15.2)
Considering PPP, let N (t) denote a counting function that represents the number
of packet arrivals in [0, t]. The packets are assumed to arrive with a rate λ, which,
in the scope of this book, is assumed to be a constant, and thus the PPP hereafter is
regarded as homogeneous or stationary PPP unless stated otherwise. As defined in
Definition 9.3, the packet arrivals in the PPP occur one at a time, N (t) has stationary
and independent increments, and N (t) ∼ Poi(λt) with the following pmf
e−λt (λt)n
Pr [N (t) = n] = ,n ∈ N (15.3)
n!
Similar to Poisson random variable, N (t) has equal expected value and variance which
are given by
E[N (t)] = V [N (t)] = λt. (15.4)
Due to the stationary increment, the number of packet arrivals in the interval
(t1 , t2 ], t1 ≤ t2 , denoted by N (t1 , t2 ], also follows a Poisson distribution with the
following pmf:
e−λ(t2 −t1 ) (λ(t2 − t1 ))n
Pr [N (t1 , t2 ] = n] = , n ∈ N. (15.5)
n!
The average number of events in (t1 , t2 ] is thus given by
E[N (t1 , t2 ]] = λ(t2 − t1 ). (15.6)
15.4 Performance measures in queuing systems

A queuing system is characterized by various performance measures which are related
to probabilistic characteristics of flows and services in the system. This section will
briefly introduce popular performance measures in queuing systems, including inter-

arrival, service time, waiting time, response time, the number of packets in the system,
queue length, traffic intensity, and server utilization. In addition, Little’s law will be
presented to describe the relationship between these measures. Note that the notation
defined in this section will be used hereafter throughout the remaining chapters.
15.4.1 Interarrival and service time in Poisson point process

Definition 15.1. Interarrival and service time are defined as follows:
● Interarrival time is the time difference between packet arrivals.
● Service time is the time difference between service completions.
Let TI denote the interarrival time between packets in PPP. As shown in Theorem
9.1, TI is exponentially distributed with an arrival rate of λ, i.e., TI ∼ Exp(λ) (cf.
Section 13.3 in Chapter 13 for details of exponential random variable). The probability
density function (pdf), i.e., fTI (t), and cumulative distribution function (cdf), i.e.,
FTI (t), of TI are given by Definition 13.2 and Theorem 13.4, respectively, i.e.,
fTI (t) = λe−λt , (15.7)
−λt
FTI (t) = 1 − e . (15.8)
The expected value and variance of TI are determined in Theorem 13.5 as
1
E[TI ] = , (15.9)
λ
1
V [TI ] = 2 . (15.10)
λ
Similarly, when the service time, denoted by TS , is exponentially distributed with
a service rate μ, i.e., TS ∼ Exp(μ), the pdf, cdf, expected value, and variance of TS
are, respectively, given by
fTS (t) = λe−μt , (15.11)
−μt
FTS (t) = 1 − e , (15.12)
1
E[TS ] = , (15.13)
μ
1
V [TS ] = . (15.14)
μ2
15.4.2 Waiting and response time

Definition 15.2. The waiting time and response time are defined as follows:
● Waiting time is the duration the packets stay in the queue.
● Response time is the total time the packets spend in the system.
It can be noticed that the response time involves the waiting time in the queue
and the service time at the server.
Let TW and TR denote the waiting time and response time, respectively. We then
have the following relationship:
TR = TW + TS , (15.15)
where TS is the service time defined in Definition 15.1.
15.4.3 The number of packets in the queue and in the system

Definition 15.3. The number of packets in the queue and in the system is defined as
follows:
● The number of packets in the queue, a.k.a. queue length, is the number of packets
waiting in the queue to be served.
● The number of packets in the system is the total number of packets staying in the
system including those being served.
Note that the number of packets being served is the number of servers in the
system.
Let ns , NQ (t), and N (t) denote the number of servers, the number of packets in
the queue, and the number of packets in the system, respectively. We then have
NQ (t) = [N (t) − ns ]+ , (15.16)
where [a]+ max{a, 0}.
15.4.4 Traffic intensity and server utilization

Traffic intensity and server utilization are defined as follows:
Definition 15.4. Traffic intensity, denoted by ρ, is defined as a measure of the ability

of a server to serve requests, which is determined by
E[TS ]
ρ . (15.17)
E[TI ]
Remark 15.1. When both TS and TI are exponential distributed with E[TS ] = 1/μ
and E[TI ] = 1/λ, we have
λ
ρ= . (15.18)
μ
Definition 15.5. Server utilization, denoted by US , is defined as a percentage of time

that the server is busy serving requests, i.e.,
E[TB ]
US = , (15.19)
E[TB ] + E[T0 ]
where TB and T0 are the busy period and idle period, respectively, of the server.
Remark 15.2. From (15.19), the average busy period of the server can be
determined by
US
E[TB ] = E[T0 ]. (15.20)
1 − US
Remark 15.3. It can be observed that the utilization of server is indeed the probability
that the server is busy, and thus it can be obtained by
US = 1 − P0 , (15.21)
where P0 is the steady-state probability that the server is idle. Substituting (15.21)
into (15.19), P0 can be derived as
E[T0 ]
P0 = . (15.22)
E[TB ] + E[T0 ]
Definition 15.6. The steady-state distribution of the birth–death process is repre-

sented by the following relationship of the steady-state probabilities:
n−1
λk
Pn = nk=0 P0 , (15.23)
k=1 μk
where Pn is the steady-state probability that there are n packets in the system, λk is
the arrival rate of kth packet, and μk is the service rate of the kth packet.
Theorem 15.1. The steady-state probability that the system is idle can be given by
∞ n−1
−1
λ k
P0 = 1 + nk=0 . (15.24)
n=1 k=1 μk
Proof. The proof follows from the fact that

∞
Pn = 1
n=0
n−1
(a)
k=0 λk
∞
⇔ P0 + n P0 = 1
n=1 k=1 μk
n−1 −1
∞ λk
⇔ P0 = 1 + nk=0 ,
n=1 μ
k=1 k
15.4.5 Little’s law for queuing systems

Figure 15.2 illustrates a simple queuing model with fundamental and explicit
performance measures. Their relationship is shown in the following Little’s law.
TR, N
TW, NQ
λ, TI μ, TS
Figure 15.2 Performance measures in a queuing system
Definition 15.7. Little’s law is used to show the relationship between performance
measures in queuing systems. We have the following relations:
E[N (t)] = λE[TR ], (15.25)
E[NQ (t)] = λE[TW ], (15.26)
E[TR ] = E[TW ] + E[TS ]. (15.27)
Theorem 15.2. The traffic intensity represents the difference between the average
number of packets in the queue and those in the system, i.e.,
ρ = E[N (t)] − E[NQ (t)]. (15.28)
Proof. Substituting (15.27) into (15.25) with (15.26), we obtain
E[N (t)] = E[NQ (t)] + λE[TS ]

(a) λ
= E[NQ (t)] +
μ
(b)
= E[NQ (t)] + ρ,
where (a) follows from (15.13), i.e., E[TS ] = 1/μ, and (b) follows from Definition
15.4 of traffic intensity, i.e., ρ = λ/μ. This completes the proof.
15.5 Kendall’s notation for queuing systems
Kendall’s notation is used to describe queuing systems via six parameters represented
as follows:
A/B/ns /nk /np /S,
where
● A denotes the distribution of packet interarrivals.
● B denotes the distribution of service time at the server. For communication net-
works, a server is simply a communication link. In this case, the link transmission
time can be regarded as the service time.
● ns is the number of servers in the system. For communication networks, ns means

the number of communication links between nodes.
● nk is the system capacity which is the maximum number of packets in the system
including all packets in the queue and those being served.
● np is the population size, which is defined as the number of sources of arriving
packets.
● S denotes the queuing discipline which specifies the order or priorities of packet
arrivals at the queue and the packets going out for service. Some commonly used
disciplines are FIFO or FCFS, LIFO, FIRO, etc.
In particular, for the first two letters A and B in the Kendall’s notation, they are
commonly described by one of the following symbols:
● M : means exponential distribution as in Markovian process.
● G: means general distribution which can be defined by either pdf, cdf, or arbitrary
distribution.
● D: means deterministic or constant in which all packet interarrivals and service
time are invariant.
In the Kendall’s notation, if the system capacity is infinity, i.e., nk → ∞, the
population size is infinity, i.e., np → ∞, and the service discipline is FIFO, i.e., S =
FIFO, then they can be omitted for simplicity.
Example 15.1. We have some queuing models as follows:
● M /M /1 queue describes a single-server queue with Poisson arrivals, exponential
service time distribution, infinite system capacity, infinity population, and FIFO
discipline.
● M /G/c/c queue describes a c-server queue with Poisson arrivals, generally
distributed service time, finite system capacity with a maximum of c packets in
the system, infinite population, and FIFO discipline.
15.6 Related works

references:
Queuing Theory Takacs in 1962 [268], Newell in 1971 [269], Nelson in 1995 [220],
and Harchol-Balter in 2013 [178].
Queuing Systems Kleinrock in 1975 and 1976 [270,271], Giambene in 2005 [272],
and Alfa in 2010 [273].
PPP for Queuing Systems Wolff in 1982 [274], Van Doorn and Regterschot in 1988
[275], and Nelson in 1995 [220].
Queuing Performance Measures Little in 1961 [276], Stidham in 1974 [277],
Nelson in 1995 [220], Stewart in 2009 [246], and Harchol-Balter in 2013 [178].
Kendall’s Notation Kendall in 1954 [278], Lee in 1966 [279], and Kleinrock in
1975 [270].

[Q15.1] Provide some applications of queuing theory in telecommunications.
[Q15.2] What are the components of a queuing system in computer and commu-
nication networks?
[Q15.3] Describe queue capacity and discipline in a queuing system.
[Q15.4] What is birth–death process?
[Q15.5] How to represent a counting process with a Poisson point process (PPP)?
[Q15.6] Describe the probability mass function (pmf) of a Poisson random variable
X with arrival rate λ.
[Q15.7] Derive the pmf of a PPP N (t) over time.
[Q15.8] What are interarrival and service time in a queuing system?
[Q15.9] Derive probability density function (pdf) and cumulative distribution
function (cdf) of the interarrival and service time in PPP.
[Q15.10] What are waiting time and response time?
[Q15.11] What are traffic intensity and server utilization?
[Q15.12] What is steady-state distribution in a birth–death process?
[Q15.13] Derive the steady-state probability that the system is idle.
[Q15.14] What is Little’s law for queuing systems?
[Q15.15] Provide an illustration to show the relationship between performance
measures in a simple single-server queuing system.
[Q15.16] Describe Kendall’s notation and explain all parameters which are used to
represent a queuing system.
15.8 Problems
[P15.1] Show the relationship between the number of packets in the queue and those
in the system.
[P15.2] Derive the average busy and idle periods of the server as a function of the
server utilization.
[P15.3] Derive the traffic intensity using Little’s law.
[P15.4] Interpret the following Kendall’s notation for various queuing systems:
● M /M /1/K
● M /M /c/c
● M /M /∞
● M /M /c/K/L
● M /G/1/K/L
● D/D/c/K/K
[P15.5] A PPP N (t) in a queuing system has an arrival rate λ = 4 packets/second.
● Derive the pmf and cdf of N (t).
● Find the expected value of N (t) during an interval of a minute.
● Find the probability that there are 100 packets arrived in a minute.
[P15.6] A single-server queuing system has both interarrival time and service time
following exponential distribution with an average arrival rate λ = 10
packets/second and a service rate μ = 12 packets/second, respectively.
● Give the Kendall’s notation describing this system.
● Derive the pmf and cdf of both the interarrival time and service time.
● Find the traffic intensity of the system.
Chapter 16
Single-server queues—network behaviors
and analysis
Queuing models can be generally classified based on the number of servers in the
system. This chapter is devoted to present single-server queues in which packets are
sequentially processed by only one server. If there are more than one packet arriving,
then they have to wait until the service at the server is complete. Depending on
the system capacity and population size of the sources, various models have been
developed for single-server queues.
In this chapter, I will first introduce M /M /1 queue with infinite system capacity
and infinite population in Section 16.1. Variant models of M /M /1 queue with finite
system capacity and finite population will be sequentially discussed in Sections 16.2
and 16.3. Specifically, network behaviors and their relevant performance measures
will be provided in detail for every queuing model.
16.1 M /M /1 Queue
In M /M /1 queue, there is only a server. The system capacity and the population,
which by default are not given, mean infinite (cf. Figure 16.1). Both interarrival time,
i.e., TI , and service time, i.e., TS , are exponentially distributed with E[TI ] = 1/λ and
E[TS ] = 1/μ.
∞
Infinite source Queue Server
FIFO
Infinite system capacity
Figure 16.1 M /M /1 queue

268 Network design, modeling and performance evaluation
Remark 16.1. In single-server queues, in order to prevent from the overload of the
queue, it is required that the arrival rate should be less than the service rate, i.e.,
λ < μ. In other terms, the traffic intensity ρ should be
λ
ρ= < 1. (16.1)
μ
Theorem 16.1. The steady-state probability that there are n packets, n = 0, 1, 2, . . . ,

in the M /M /1 queue is given by
Pn = (1 − ρ)ρ n . (16.2)

n−1 −1
λk
∞
P0 = 1 + nk=0
n=1
μ
k=1 k
∞ n −1

(a) λ
= 1+
n=1
μ
∞ −1

= ρn
n=0
(b)
= 1−ρ
where (a) follows from the assumption that λk = λ and μk = μ ∀k and (b) follows
from the infinite sum of geometric series, i.e.,

∞
1
qx = , ∀|q| < 1.
x=0
1−q
Substituting P0 into (15.23), we obtain

n−1
λ
P n = P0
k=0
μ
= (1 − ρ)ρ n .
Theorem 16.2. The average number of packets in the system and in the queue of
M /M /1 system is, respectively, given by
ρ
E[N ] = , (16.3)
1−ρ
ρ2
E[NQ ] = . (16.4)
1−ρ
Single-server queues—network behaviors and analysis 269
Proof. We have
∞
E[N ] = nPn
n=0
(a)
∞
= n(1 − ρ)ρ n
n=0

∞
= (1 − ρ)ρ nρ n−1
n=1
(b) 1
= (1 − ρ)ρ
(1 − ρ)2
ρ
= ,
1−ρ
where (a) follows from (16.2) and (b) follows from the derivative of geometric
progression, i.e.,
∞
1
nxn−1 = , ∀|x| < 1.
x=1
(1 − x)2
From (15.28) in Theorem 15.2, we have
E[NQ ] = E[N ] − ρ
(a) ρ
= −ρ
1−ρ
ρ2
= ,
1−ρ
Theorem 16.3. The average response time and waiting time in M /M /1 queue are,
1
E[TR ] = , (16.5)
μ−λ
ρ
E[TW ] = . (16.6)
μ−λ
Proof. From the Little’s law in Section 15.4.5, we have

E[N ]
E[TR ] =
λ
(a) (ρ/(1 − ρ))
=
λ
1
= .
μ−λ
We also have
E[TW ] = E[TR ] − E[TS ]

1 1
= −
μ−λ μ
ρ
= .
μ−λ
Theorem 16.4. Server utilization and the average busy period of the server in M /M /1
queue are, respectively, given by
US = ρ, (16.7)
1
E[TB ] = . (16.8)
μ−λ
Proof. From (15.21), we have
US = 1 − P 0
= 1 − (1 − ρ)
= ρ,
where the second equal sign follows from (16.2) in Theorem 16.1 with n = 0, i.e.,
P0 = 1 − ρ.
From (15.22) with a notice that the average idle time of the server, i.e., E[T0 ] =
1/λ, we have
E[T0 ]
P0 =
E[TB ] + E[T0 ]
1/λ
= .
E[TB ] + (1/λ)
Therefore, we can obtain E[TB ] as
1/λ 1
E[TB ] = −
1−ρ λ
(ρ/(1 − ρ))
=
λ
1
= .
μ−λ
16.2 M /M /1/K Queue

An M /M /1/K queue is an extended M /M /1 queue with a finite capacity of K packets,
i.e., maximum K packets, in the system (cf. Figure 16.2). Both interarrival time, i.e.,
TI , and service time, i.e., TS , are also exponentially distributed with E[TI ] = 1/λ and
E[TS ] = 1/μ.
∞
Infinite source Queue Server
FIFO
System capacity = K
Figure 16.2 M /M /1/K queue
Theorem 16.5. The steady-state probability that there are n packets, n =

0, 1, 2, . . . , K, in M /M /1/K queue is given by
⎧
⎪ 1
⎪
⎨K + 1 if ρ = 1,
ρ n
Pn = K = (16.9)
k=0 ρ
k ⎪
⎪ 1−ρ n
⎩ ρ if 0 ≤ ρ < 1.
1 − ρ K+1

n−1 −1

K
λk
P0 = 1+ nk=0
n=1 k=1 μk
K n
−1
(a) λ
= 1+
n=1
μ
K −1

= ρ n
n=0
⎧
1⎪
⎨ if ρ = 1,
= K 1+−1ρ
(b)
⎪
⎩ if 0 ≤ ρ < 1,
1 − ρ K+1
where (a) follows from the assumption that λk = λ and μk = μ ∀k and (b) follows
from the infinite sum of geometric series, i.e.,

K
1 − qK+1
qx = , ∀|q| < 1.
x=0
1−q

n−1
λ
P n = P0
k=0
μ
ρn
= K
k=0 ρk
⎧
⎪ 1
⎨ if ρ = 1,
= K 1+−1ρ
⎪
⎩ ρn if ρ < 1.
1 − ρ K+1
Theorem 16.6. The average number of packets in the system and in the queue of the
M /M /1/K system is, respectively, given by

K
kρ k
E[N ] = K , (16.10)
k=1 n=0 ρn
−1

K
kρ k
K
E[NQ ] = K + ρ n
− 1. (16.11)
k=1 n=0 ρn n=0
Proof. We have

K
E[N ] = kPk
k=0
(a)
K
kρ k
= K
k=0 n=0 ρn

K
kρ k
= K ,
k=1 n=0 ρn

We also have

K
E[NQ ] = (k − 1)Pk
k=1

K
K
= kPk − Pk
k=1 k=1
(a)
= E[N ] − (1 − P0 )
K −1
(b)
K
kρ k
= K + ρ n
−1
n=0 ρ
n
k=1 n=0
K
where (a) follows from k=0 Pk = 1 and (b) follows from (16.10) and (16.9).
Remark 16.2. In an M /M /1/K queue, due to its limited capacity, the system is
blocked when there are K packets in the system. In other words, the (K + 1)th packet
will be dropped. The blocking probability of the queuing system can be understood
as the probability that the system is full, which is given by
ρK
P K = K . (16.12)
k=0 ρk
It is also noticed that the packet arrival only occurs when the system is unblocked.
Let λ denote the average arrival rate into the system. We then have
λ = λ(1 − PK )

ρK
= λ 1 − K (16.13)
k=0 ρk
Remark 16.3. Another observation from (16.10), (16.11), and (16.13) is that
E[N ] = E[NQ ] + ρ(1 − PK ) (16.14)
Theorem 16.7. The average response time and waiting time in M /M /1/K queue are,
K
(kρ k / Kn=0 ρ n )
E[TR ] = k=1 , (16.15)
λ 1 − (ρ K / Kk=0 ρ k )
K K −1
K
k=1 (kρ k
/ n=0 ρ n
) + n=0 ρ n
−1
E[TW ] = . (16.16)
λ 1 − (ρ K / Kk=0 ρ k )
Proof. The proof follows from the Little’s law in Section 15.4.5 with
E[N ]
E[TR ] = ,
λ
E[NQ ]
E[TW ] = ,
λ
where λ is given by (16.13), and E[N ] and E[NQ ] are given by (16.10) and (16.11),
respectively.
Theorem 16.8. Server utilization and the average busy period of the server in
M /M /1/K queue are, respectively, given by
K −1

US = 1 − ρn , (16.17)
n=0
K
1 n
E[TB ] = ρ −1 . (16.18)
λ n=0
Proof. The proof follows from (15.21) and (15.20), i.e.,

US = 1 − P0 ,
US
E[TB ] = E[T0 ],
1 − US
where P0 is given by (16.9) and E[T0 ] = 1/λ.
16.3 M/M/1/K/K Queue

An M /M /1/K/K queue is a generalized version of M /M /1/K queue with a finite
source having population size of K (cf. Figure 16.3). All the other parameters are
similar to those in the M /M /1/K queuing system with the system capacity K, expo-
nentially distributed interarrival time, i.e., TI , with E[TI ] = 1/λ, and exponentially
distributed service time, i.e., TS , with E[TS ] = 1/μ.
Remark 16.4. With a finite source, it can be noticed that the arrival rate at the nth
stage is reduced as the birth rate in the birth–death process with

(K − n)λ if 0 ≤ n ≤ K,
λn = (16.19)
0 if n > K,
while the service rate is a constant death rate due to the fact that there is only a single
server in the system, i.e.,
μn = μ, 1 ≤ n ≤ K. (16.20)
Queue Server
FIFO
System capacity = K
K sources
Figure 16.3 M /M /1/K/K queue

0, 1, 2, . . . , K, in M /M /1/K/K queue is given by
(K!/(K − n)!)ρ n
Pn = K . (16.21)
k=0 (K!/(K − k)!)ρ
k
n−1 −1

K
λk
P0 = 1+ nk=0
n=1 k=1 μk
n −1
(a)
K
n−1
λ
= 1+ (K − k)
n=1 k=0
μ
−1
(b) K
K!
= 1+ ρ n
n=1
(K − n)!
K −1
K!
= ρ n
,
n=0
(K − n)!
where (a) follows from (16.19), (16.20), and ρ = λ/μ, and (b) follows from the fact
that

n−1
K!
(K − k) = .
k=0
(K − n)!

n−1 n
λ
P n = P0 (K − k)
k=0
μ
(K!/(K − n)!)ρ n
= K .
k=0 (K!/(K − k)!)ρ
k
the M /M /1/K/K system is, respectively, given by
⎛ K −1 ⎞
1 K!
E[N ] = K − ⎝1 − ρn ⎠, (16.22)
ρ n=0
(K − n)!
⎛ K −1 ⎞
K!
1 ⎝ ⎠.
E[NQ ] = K − 1 + 1− ρn (16.23)
ρ n=0
(K − n)!
Proof. We have

K
E[N ] = kPk
k=0

K
= (KPk − (K − k)Pk )
k=0
(a)
K
= K− (K − k)Pk
k=0
1
K−1
(b)
= K− Pk+1
ρ k=0
(c) 1
= K− (1 − P0 )
ρ
⎛ K −1 ⎞
1⎝ K!
(d)
= K− 1− ρn ⎠,
ρ n=0
(K − n)!

where (a) and (c) follow from Kk=0 Pk = 1, (b) follows from the fact that Pk+1 =
ρ(K − k)Pk , and (d) follows from (16.21).
We also have

K
E[NQ ] = (k − 1)Pk
k=1

K
K
= kPk − Pk
k=1 k=1
(a)
= E[N ] − (1 − P0 )
(b) 1
= K − (1 − P0 ) − (1 − P0 )
ρ

1
= K − 1+ (1 − P0 )
ρ
⎛ K −1 ⎞
K!
(c) 1 ⎝ ⎠
= K − 1+ 1− ρn
ρ n=0
(K − n)!

where (a) follows from Kk=0 Pk = 1, (b) follows from the derivation of (16.22), and
(c) follows from (16.21).
Remark 16.5. Notice that in an M /M /1/K/K queue, apart from the packets in the
system, there are packets waiting in the source. Let NS denote the number of packets
in the source. The average number of packets in the source can be given by
E[NS ] = K − E[N ]
⎛ K −1 ⎞
1 K!
= ⎝1 − ρn ⎠. (16.24)
ρ n=0
(K − n)!
The utilization of the source, denoted by USource , is thus given by

E[NS ]
USource =
K
⎛ K −1 ⎞
1 ⎝ K!
= 1− ρn ⎠. (16.25)
Kρ n=0
(K − n)!
Theorem 16.11. The average response time and waiting time in M /M /1/K/K queue
−1
K
K − (1/ρ) 1 − n=0 (K!/(K − n)!)ρ
n
E[TR ] = −1 , (16.26)

K
μ 1− n=0 (K!/(K − n)!)ρ n
−1
K
K − (1 + (1/ρ)) 1 − n=0 (K!/(K − n)!)ρ
n
E[TW ] = −1 . (16.27)

K
μ 1− n=0 (K!/(K − n)!)ρ
n
E[N ]
E[TR ] = ,
λE[NS ]
E[NQ ]
E[TW ] = ,
λE[NS ]
where E[NS ], E[N ], and E[NQ ] are given by (16.24), (16.22), and (16.23),
respectively.
M /M /1/K/K queue are, respectively, given by
K −1
K!
US = 1 − ρ n
, (16.28)
n=0
(K − n)!
K
1 K!
E[TB ] = ρ −1 .
n
(16.29)
Kλ n=0
(K − n)!
Proof. The proof is similar to that for Theorem 16.8 by using (15.21) and (15.20),
i.e.,
US = 1 − P 0 ,
US
E[TB ] = E[T0 ],
1 − US
with P0 given by (16.21) and the average idle time of the server E[T0 ] = 1/(Kλ).
16.4 Related works

references: Takacs in 1962 and 1963 [268,280], Chang in 1970 [281], Kleinrock in
1975 [270], Gopinath and Morrison in 1977 [282], Eckberg in 1979 [283], Wang and
Mandayam in 2004 [284], Stewart in 2009 [246], Steiner and Shamai in 2010 [285],
Cohen in 2012 [286], and Zhao et al. in 2018 [287].

[Q16.1] Describe the following single-server queuing systems: M /M /1,
M /M /1/K and M /M /1/K/K. Provide an illustrative example for each
of them.
[Q16.2] What is the stability condition for a single-server queue?
[Q16.3] Derive the steady-state probability that there are n packets in M /M /1,
M /M /1/K and M /M /1/K/K queues.
[Q16.4] Derive the average number of packets in the system and in the queue of
M /M /1, M /M /1/K and M /M /1/K/K queues.
[Q16.5] Derive the average response time and waiting time in M /M /1, M /M /1/K,
and M /M /1/K/K queues.
[Q16.6] Derive server utilization and the average busy period of the server in
M /M /1, M /M /1/K, and M /M /1/K/K queues.
[Q16.7] When is a queuing system blocked?
[Q16.8] Derive the blocking probability of an M /M /1/K queue.
[Q16.9] Derive the average number of packets in the source and the utilization of
the source in an M /M /1/K/K queue.
16.6 Problems
[P16.1] Consider a server in a network. Packets arrive on an average of 5 s and it
takes an average of 4 s for processing at the server. Both the inter-arrival
time and service time are exponentially distributed. The queue is assumed
to be infinitely large (unbounded) and the serving discipline is FIFO.
● Give the Kendall’s notation describing this system and check the
condition for stability of this system.
● Find the probability that the server is busy.
● Find the probability that there are less than three packets in the system.
● Find the average number of packets waiting to be processed and the
average number of packets in the system.
● Find the average time the packet spends in the queue and totally spends
in the system.
[P16.2] An M/M/1 queue has a service rate of two customers per hour.
● Find the maximum arrival rate such that the average queuing interval
is not longer than 30 min.
● Using the obtained maximum arrival rate,
– Find the average number of customers in the queue and in the
system.
– Find the probability that the server is busy.
– Find the probability that there are no more than two customers in
the system.
[P16.3] The data arrivals at a server are assumed to follow Poisson point process.
The time to process data at the server is exponentially distributed with a
mean of 1.5 s. The average waiting time of the packet in the queue is 3 s.
● Find the arrival rate of the packets.
● Find the server utilization and the average busy period of the server.
● Find the average number of packets in the queue and in the system.
● Find the probability that there is at least a packet in the system.
[P16.4] At a database server, it takes an average of 25 s to store a particular data
packet. The packet arrival and storing time at the server follow exponential
distribution.
● Find the average interarrival time of the data packets given that the
average queuing length is five data packets.
● Find the average response and waiting time in the system.
[P16.5] A single-server queue has a finite capacity of five packets. Both the inter-
arrival and service time are exponentially distributed with the average
interarrival of 10 s and the average service time of 8 s.
● Find the probability that the server is idle and the blocking probability
of the queuing system.
● Find the average number of packets in the system and in the queue.
● Find the average response time and waiting time of the data packets.
● Find the server utilization and the average busy period of the server.
[P16.6] Considering a finite source in a single-server queuing system with a limited
capacity of five packets and a population size of five packets. Both the
packet arrival and service follow Poisson point process with an arrival rate
of 5 packets/s and a service rate of 8 packets/s.
● Find the probability that the server is busy.
● Find the average number of packets in the system, in the queue and in
the source.
● Find the server utilization, source utilization and the average busy
period of the server.
Chapter 17
Multi-server queues—network behaviors and
analysis
Considering queuing systems with multiple servers, this chapter is devoted to present
multi-server queues in which packets are served by multiple packets in parallel. There
exist various models for multi-server queues depending on the system capacity and
population size of the sources.
In this chapter, I will first introduce M /M /c queue with infinite system capacity
and infinite population in Section 17.1. Its variant models with finite system capacity
and finite population will be sequentially discussed in Sections 17.2 and 17.3. For
every queuing model, its behaviors will be presented along with detailed analyses of
relevant performance measures.
17.1 M /M /c Queue
In M /M /c queue, there are c servers. By default, the capacity and population without
stated means infinite (cf. Figure 17.1). Both interarrival time, i.e., TI , and service
time, i.e., TS , are exponentially distributed with E[TI ] = 1/λ and E[TS ] = 1/μ,
respectively.
Remark 17.1. With multiple servers to serve in parallel, it can be noticed that the
service rate at the nth stage increases as the death rate in the birth–death process
with
⎧
⎨nμ if 0 ≤ n ≤ c,
μn = (17.1)
⎩cμ if n > c,
where the second case is due to the maximum service rate with c servers is cμ. The
arrival rate is however a constant birth rate, i.e.,
λn = λ, n ≥ 0. (17.2)
Remark 17.2. In multi-server queues, in order to prevent from the overload of the
queue, it is required that the arrival rate should be less than the service rate, i.e.,
λ < cμ. In other terms, the traffic intensity, i.e., ρ = λ/μ, should be
ρ < c. (17.3)
∞ ...
...
Infinite source Queue
FIFO
c Servers
Infinite system capacity
Figure 17.1 M /M /c Queue
For simplicity, we define an auxiliary parameter

ρ
a . (17.4)
c
Theorem 17.1. The steady-state probability that there are n packets, n = 0, 1, 2, . . . ,

in the M /M /c queue is given by
⎧ −1
⎪
⎪ ρk
ρ n c−1 ρc
⎪
⎨ + if 0 ≤ n ≤ c,
n! k=0 k! c!(1 − a)
Pn = −1 (17.5)
⎪
⎪ ρk
an cc c−1 ρc
⎪
⎩ + if n > c.
c! k=0 k! c!(1 − a)
n−1 −1

∞
λk
P0 = 1+ nk=0
n=1 k=1 μk
−1
(a)
c−1
λn
∞
λn
= 1+ n
+
n=1
n!μ n=c
c!c (μ)n
n−c
c−1 −1

ρn ρ c
ρ n−c
∞
= +
n=0
n! c! n=c cn−c
Multi-server queues—network behaviors and analysis 283
c−1 −1
(b)
ρn ρc
n
∞
= + a
n=0
n! c! n=0
c−1 −1
(c)
ρn ρc
= + ,
n=0
n! c!(1 − a)
where (a) follows from (17.1) and (17.2), (b) is due to a = ρ/c, and (c) follows from
the infinite sum of geometric series, i.e.,
∞
1
qx = , ∀|q| < 1.
x=0
1−q
n−1
λk
Pn = P0 nk=0
k=1 kμ
⎧ λn
⎪
⎨P 0 if 0 ≤ n ≤ c,
= n!μn
⎪ λ n
⎩P0 if n > c
c!c μn
n−c
⎧
⎪ ρn
⎨P0 if 0 ≤ n ≤ c,
= n!
n c
⎪
⎩P0 a c if n > c
c!
⎧ −1
⎪
⎪ ρk
ρ n c−1 ρc
⎪
⎨ + if 0 ≤ n ≤ c,
n! k=0 k! c!(1 − a)
= c−1 k −1
⎪
⎪ an c
c ρ ρc
⎪
⎩ + if n > c.
c! k=0 k! c!(1 − a)
Remark 17.3. Notice that, if the number of arriving packets is larger than the number
of servers, then only c packets can be served in parallel with c servers, while the
remaining packets have to wait.
Theorem 17.2. The waiting probability of a packet, i.e., PW , in M /M /c queue is

given by
c−1 −1

∞
an c c
ρ k ρc
PW = + . (17.6)
n=c
c! k=0
k! c!(1 − a)
Proof. By definition, the waiting probability of a packet is corresponding to the

probability that the system already has at least c packets, i.e.,
PW = Pn≥c

∞
= Pn
n=c
c−1 −1

∞
an c c
ρk ρc
= + ,
n=c
c! k=0
k! c!(1 − a)
where the last equal sign follows from (17.5).
M /M /c system are, respectively, given by
c−1 −1

∞
an c c
ρ k ρc
E[N ] = ρ + (n − c) + , (17.7)
n=c+1
c! k=0
k! c!(1 − a)
c−1 −1

∞
an c c
ρk ρc
E[NQ ] = (n − c) + . (17.8)
n=c+1
c! k=0
k! c!(1 − a)
Proof. Let us first derive E[NQ ]. We have
∞
E[NQ ] = (n − c)Pn
n=c+1
c−1 −1
(a)
∞
an c c
ρk ρc
= (n − c) + ,
n=c+1
c! k=0
k! c!(1 − a)

From (15.28) in Theorem 15.2, we have
E[N ] = E[NQ ] + ρ
c−1 −1

∞
an c c
ρk ρc
= ρ+ (n − c) + .
n=c+1
c! k=0
k! c!(1 − a)
Theorem 17.4. The average response time and waiting time in M /M /c queue are,
⎡ c−1 −1 ⎤
1⎣
∞
an c c
ρ k ρc
E[TR ] = ρ+ (n − c) + ⎦, (17.9)
λ n=c
c! k=0
k! c!(1 − a)
c−1 −1
1

ρk
∞
an c c ρc
E[TW ] = (n − c) + . (17.10)
λ n=c c! k=0
k! c!(1 − a)
Proof. From the Little’s law in Section 15.4.5, we have

E[N ]
E[TR ] =
λ
⎡ c−1 −1 ⎤
(a) 1
∞
a n c
c
ρk ρ c
= ⎣ρ + (n − c) + ⎦.
λ n=c
c! k=0
k! c!(1 − a)

We also have
E[NQ ]
E[TW ] =
λ
c−1 −1
1

ρk
∞
(a) an c c ρc
= (n − c) + ,
λ n=c c! k=0
k! c!(1 − a)
Remark 17.4. Note that the traffic intensity in multi-server queue is in fact the
utilization of all servers. Let UA denote the utilization of all servers. We then have
UA = ρ. (17.11)
This overall utilization also means the average number of busy servers in the system.
Let cB and cI denote the number of busy servers and the number of idle servers,
respectively. We can arrive at
E[cB ] = ρ, (17.12)
E[cI ] = c − ρ. (17.13)
Remark 17.5. Considering the whole system, the system utilization in the multi-
server queue, denoted by USystem , is similarly defined as in (15.21), i.e.,
USystem = 1 − P0
c−1 −1

ρn ρc
= 1− + , (17.14)
n=0
n! c!(1 − a)
where P0 is given by (17.5).
Theorem 17.5. Server utilization and the average busy period of the server in M /M /c
queue are, respectively, given by
ρ
US = a = , (17.15)
c
c−1
1
ρn ρc
E[TB ] = + −1 . (17.16)
λ n=0
n! c!(1 − a)
Proof. We have
UA = cUS ,
and thus
UA
US =
c
ρ
= = a.
c
From (15.22) with a notice that the average idle time of the server, i.e.,
E[T0 ] = 1/λ, we have
E[T0 ]
P0 =
E[TB ] + E[T0 ]
1/λ
= .
E[TB ] + (1/λ)
Therefore, we can obtain E[TB ] as
1 − P0
E[TB ] =
λP0
c−1
(a) 1
ρn ρc
= + −1 ,
λ n=0
n! c!(1 − a)
where (a) follows from (17.5) with n = 0.
17.2 M /M /c/K Queue
An M /M /c/K queue is an extended version of M /M /c queue for the case of finite

capacity with a maximum of K packets in the system (cf. Figure 17.2). Both interar-
rival time, i.e., TI , and service time, i.e., TS , are also exponentially distributed with
E[TI ] = 1/λ and E[TS ] = 1/μ.
Theorem 17.6. The steady-state probability that there are n packets, n = 0, 1,

2, . . . , K, in the M /M /c/K queue is given by
⎧ −1
⎪
⎪ ρk
ρ n c−1 ρ c (1 − aK−c+1 )
⎪
⎨ + if 0 ≤ n ≤ c,
n! k=0 k! c!(1 − a)
Pn = −1 (17.17)
⎪
⎪ ρk
an cc c−1 ρ c (1 − aK−c+1 )
⎪
⎩ + if c < n ≤ K,
c! k=0 k! c!(1 − a)
where a (ρ/c) < 1.
...
Infinite source Queue
FIFO
c Servers
System capacity = K
Figure 17.2 M /M /c/K Queue
n−1 −1

K
λk
P0 = 1+ nk=0
n=1 k=1 kμ
−1
(a)
c−1
λn
K
λn
= 1+ n
+
n=1
n!μ n=c
c!c (μ)n
n−c
c−1 −1

ρn ρ c
ρ n−c
K
= +
n=0
n! c! n=c cn−c
c−1 −1

ρn ρc
n
K−c
= + a
n=0
n! c! n=0
c−1 −1
(b)
ρn ρ c (1 − aK−c+1 )
= + ,
n=0
n! c!(1 − a)
where (a) follows from (17.1) and (17.2), and (b) follows from
K
1 − aK−c+1
ak−c = if a = 1. (17.18)
k=c
1−a

n−1
λk
Pn = P0 nk=0
k=1 μk
⎧ λn
⎪
⎨P0 if 0 ≤ n ≤ c,
= n!μn
⎪ λ n
⎩P0 if c < n ≤ K
c!c μn
n−c
⎧
⎪ ρn
⎨P0 if 0 ≤ n ≤ c,
= n!
n c
⎪
⎩P0 a c if c < n ≤ K
c!
⎧ −1
⎪
⎪ ρk
ρ n c−1 ρ c (1 − aK−c+1 )
⎪
⎨ + if 0 ≤ n ≤ c,
n! k=0 k! c!(1 − a)
= c−1 k −1
⎪
⎪ an c
c ρ ρ c
(1 − aK−c+1 )
⎪
⎩ + if c < n ≤ K.
c! k=0 k! c!(1 − a)
Theorem 17.7. The waiting probability of a packet in M /M /c/K queue is given by

c−1 −1

K
an c c
ρ k ρ c (1 − aK−c+1 )
PW = + . (17.19)
n=c
c! k=0
k! c!(1 − a)
Proof. The waiting probability can be similarly derived as in Theorem 17.2 with a
notice that the number of packets allowed in the system is limited by K, i.e.,
PW = Pc≤n≤K

K
= Pn
n=c
c−1 −1

K
an c c
ρk ρ c (1 − aK−c+1 )
= + ,
n=c
c! k=0
k! c!(1 − a)
where the last equal sign follows from (17.17).
M /M /c/K system is, respectively, given by
⎛ c−1 −1 ⎞
a K c
c
ρk ρ c
(1 − a K−c+1
)
E[N ] = ρ ⎝1 − + ⎠
c! k=0
k! c!(1 − a)
c−1 −1

K
an c c
ρk ρ c (1 − aK−c+1 )
+ (n − c) + , (17.20)
n=c+1
c! k=0
k! c!(1 − a)
c−1 −1

K
an c c
ρk ρ c (1 − aK−c+1 )
E[NQ ] = (n − c) + . (17.21)
n=c+1
c! k=0
k! c!(1 − a)
Proof. Let us first derive E[NQ ]. We have
K
E[NQ ] = (n − c)Pn
n=c+1
c−1 −1
(a)
K
an c c
ρk ρ c (1 − aK−c+1 )
= (n − c) + ,
n=c+1
c! k=0
k! c!(1 − a)

Similar to M /M /1/K, from (16.14), we have
E[N ] = E[NQ ] + ρ(1 − PK )

⎛ c−1 −1 ⎞
K c
a c
ρ k
ρ (1 − a
c K−c+1
)
= ρ ⎝1 − + ⎠
c! k=0
k! c!(1 − a)
c−1 −1

K
an c c
ρk ρ c (1 − aK−c+1 )
+ (n − c) + .
n=c+1
c! k=0
k! c!(1 − a)
Theorem 17.9. The average response time and waiting time in M /M /c/K queue are,
1
E[TR ] =
μ
K −1
c−1
(n − c)(an cc /c!)
n=c+1 k=0 (ρ /k!) + ((ρ (1 − a
k c K−c+1
))/(c!(1 − a)))
+ −1 ,
c−1
λ 1 − (a c /c!)
K c
k=0 (ρ /k!) + ((ρ (1 − a
k c K−c+1 ))/(c!(1 − a))
(17.22)
E[TW ]
K −1
c−1
n=c+1 (n − c)(an cc /c!) k=0 (ρ /k!) + ((ρ (1 − a
k c K−c+1
))/(c!(1 − a)))
= −1 .
c−1
λ 1 − (a c /c!)
K c
k=0 (ρ /k!) + ((ρ (1 − a
k c K−c+1 ))/(c!(1 − a))
(17.23)
Proof. Let us first derive E[TW ]. Similar to the proof of Theorem 16.7, from the
Little’s law, we have
E[NQ ]
E[TW ] =
λ(1 − PK )
K −1
c−1
(a)
(n − c)(an cc /c!)
n=c+1 k=0 (ρ /k!) + ((ρ (1 − a
k c K−c+1
))/(c!(1 − a)))
= −1 ,
c−1
λ 1 − (a c /c!)
K c
k=0 (ρ /k!) + ((ρ (1 − a
k c K−c+1 ))/(c!(1 − a))
where (a) follows from (17.21) and (17.17).

We also have from the Little’s law that
E[N ]
E[TR ] =
λ(1 − PK )
(a) E[NQ ] + ρ(1 − PK )
=
λ(1 − PK )
1
= E[TW ] +
μ
(b) 1
=
μ
K −1
c−1
n=c+1 (n − c)(a c /c!) k=0 (ρ /k!) + ((ρ (1 − a ))/(c!(1 − a)))
n c k c K−c+1
+ −1 .
c−1
λ 1 − (a c /c!)
K c
k=0 (ρ /k!) + ((ρ (1 − a
k c K−c+1 ))/(c!(1 − a))
where (a) follows from (16.14) and (b) follows from (17.23).
Remark 17.6. We can similarly obtain the following measures of the M /M /c/K
queue:
● Utilization of all servers:
UA = ρ(1 − PK )
⎛ c−1 −1 ⎞
a K c
c
ρk ρ c
(1 − a K−c+1
)
= ρ ⎝1 − + ⎠. (17.24)
c! k=0
k! c!(1 − a)
● System utilization:
USystem = 1 − P0
c−1 −1

ρn ρ c (1 − aK−c+1 )
= 1− + . (17.25)
n=0
n! c!(1 − a)
● The average number of busy servers:

⎛ c−1 −1 ⎞
a K c
c
ρk ρ c
(1 − a K−c+1
)
E[cB ] = UA = ρ ⎝1 − + ⎠. (17.26)
c! k=0
k! c!(1 − a)
● The average number of idle servers:

⎛ c−1 −1 ⎞
a K c
c
ρk ρ c
(1 − a K−c+1
)
E[cI ] = c − ρ ⎝1 − + ⎠. (17.27)
c! k=0
k! c!(1 − a)
M /M /c/K queue are, respectively, given by
⎛ c−1 −1 ⎞
a K c
c
ρk ρ c
(1 − a K−c+1
)
US = a ⎝1 − + ⎠, (17.28)
c! k=0
k! c!(1 − a)
c−1
1
ρn ρ c (1 − aK−c+1 )
E[TB ] = + −1 . (17.29)
λ n=0
n! c!(1 − a)
Proof. We have
UA = cUS ,
and thus
UA
US =
c
c−1
aK c c
ρk ρ c (1 − aK−c+1 )
= a 1− + .
c! k=0
k! c!(1 − a)
1/λ, we can obtain E[TB ] as
1 − P0
E[TB ] =
λP0
c−1
1
ρn ρ c (1 − aK−c+1 )
= + −1 .
λ n=0
n! c!(1 − a)
17.3 M /M /c/K /K Queue

An M /M /c/K/K queue is a generalized version of M /M /c/K queue considering a
finite source with population size K (cf. Figure 17.3). All the other parameters are
similar to those in the M /M /c/K queuing system with c servers, system capac-
ity K, exponentially distributed interarrival time, i.e., TI , with E[TI ] = 1/λ and
exponentially distributed service time, i.e., TS , with E[TS ] = 1/μ.
...
...
Queue
FIFO
K sources
c Servers
System capacity = K
Figure 17.3 M /M /c/K/K Queue
Remark 17.7. With a finite source, it can be noticed that the arrival rate at the nth
stage is reduced as the birth rate in the birth–death process with
⎧
⎨(K − n)λ if 0 ≤ n ≤ K,
λn = (17.30)
⎩0 if n > K,
while the service rate at the nth stage increases as follows:
⎧
⎨nμ if 1 ≤ n ≤ c,
μn = (17.31)
⎩cμ if c ≤ n ≤ K.

0, 1, 2, . . . , K, in the M /M /c/K/K queue is given by
⎧ −1
⎪
⎪ K n c−1 K n ρ c K K!
⎪
⎨ n ρ ρ + a n−c
if 0 ≤ n ≤ c,
n
c! n=c (K − n)!
Pn = n=0
−1 (17.32)
⎪
⎪ K!an cc K n ρ c
c−1 K K!
⎪
⎩ ρ + a n−c
if c < n ≤ K,
(K − n)!c! n=0 n c! n=c (K − n)!
where a ρ/c.
n−1 −1

K
λk
P0 = 1+ nk=0
n=1 k=1 kμ
−1
(a)
c−1
K!λn
K
K!λn
= 1+ +
n=1
(K − n)!n!μn n=c
(K − n)!c!cn−c μn
c−1 −1
(b)
K ρ c
K
K!
= ρn + an−c ,
n=0
n c! n=c
(K − n)!

where (a) follows from (17.30),K (17.31),
and n−1
k=0 (K − n) = (K!/(K − n)!) and (b)
is due to binomial coefficient n K!/(n!(K − n)!) and a = ρ/c.
n−1
λk
Pn = P0 nk=0
k=1 μk
⎧ K!λn
⎪
⎨P 0 if 0 ≤ n ≤ c,
= (K − n)!n!μn
n
⎪
⎩P0
K!λ
if c < n ≤ K
(K − n)!c!cn−c μn
⎧
⎪
⎨P0 n ρ
K n
if 0 ≤ n ≤ c,
= n c
K!a c
⎪
⎩P0 if c < n ≤ K
(K − n)!c!
⎧ −1
⎪
⎪ K n c−1
K n ρ c K K!
⎪
⎨ n ρ ρ + a n−c
if 0 ≤ n ≤ c,
n
c! n=c (K − n)!
= n=0
−1
⎪
⎪ K!an cc K n ρ c
c−1 K K!
⎪
⎩ ρ + a n−c
if c < n ≤ K,
(K − n)!c! n=0 n c! n=c (K − n)!
Theorem 17.12. The waiting probability of a packet in M /M /c/K/K queue is

given by
c−1 −1

K
K!an cc
K ρ c
K!
K
PW = ρ +
n
a n−c
. (17.33)
n=c
(K − n)!c! n=0 n c! n=c (K − n)!

M /M /c/K/K system is, respectively, given by

c−1
−1
K n ρ c
K!
c K
K n
E[N ] = n ρ ρ + a n−c
n=0
n n=0
n c! n=c (K − n)!
c−1 −1

K
K!an cc
K ρ c
K!
K
+ n ρ +
n
a n−c
,
n=c+1
(K − n)!c! n=0 n c! n=c (K − n)!
(17.34)
c−1 −1

K
K!an cc
K ρ c
K!
K
E[NQ ] = (n − c) ρ +
n
a n−c
.
n=c+1
(K − n)!c! n=0 n c! n=c (K − n)!
(17.35)
Proof. We have
K
E[N ] = nPn
n=0

c−1
−1
K n ρ c
K!
c K
(a) K n
= n ρ ρ + a n−c
n=0
n n=0
n c! n=c (K − n)!
c−1 −1

K
K!an cc
K ρ c
K
K!
+ n ρn + an−c ,
n=c+1
(K − n)!c! n=0
n c! n=c
(K − n)!

We also have
K
E[NQ ] = (n − c)Pn
n=c+1
c−1 −1

K
K!an cc
K ρ c
K
K!
= (n − c) ρn + an−c .
n=c+1
(K − n)!c! n=0 n c! n=c (K − n)!
Remark 17.8. Notice that in an M /M /c/K/K queue, apart from the packets in the
system, there are packets waiting in the source. Let NS denote the number of packets
in the source. The average number of packets in the source can be determined by
E[NS ] = K − E[N ], (17.36)
where E[N ] is given by (17.34).
The utilization of the sources, denoted by USource , can be thus given by

E[NS ]
USource =
K
E[N ]
= 1− . (17.37)
K
Theorem 17.14. The average response time and waiting time in M /M /c/K/K queue
1 E[N ]
E[TR ] = , (17.38)
λ K − E[N ]
1 E[NQ ]
E[TW ] = , (17.39)
λ K − E[N ]
where E[N ] and E[NQ ] are given by (17.34) and (17.35), respectively.
E[N ]
E[TR ] = ,
λE[NS ]
E[NQ ]
E[TW ] = ,
λE[NS ]
where E[NS ] is given by (17.36).
Remark 17.9. We can similarly obtain the following measures of the M /M /c/K/K
queue:
● Utilization of all servers:
UA = ρE[NS ]
= ρ(K − E[N ]). (17.40)
● System utilization:
USystem = 1 − P0
c−1 −1

K ρ c
K!
K
= 1− ρ +
n
a n−c
. (17.41)
n=0
n c! n=c (K − n)!
● The average number of busy servers:
E[cB ] = UA = ρ(K − E[N ]). (17.42)
● The average number of idle servers:
E[cI ] = c − ρ(K − E[N ]). (17.43)
M /M /c/K/K queue are, respectively, given by
US = a(K − E[N ]), (17.44)
c−1
1
K ρ c
K!
K
E[TB ] = ρ +
n
a n−c
−1 . (17.45)
Kλ n=0
n c! n=c (K − n)!
Proof. We have
UA = cUS ,
and thus
UA
US =
c
= a(K − E[N ]).
1/(Kλ), we can obtain E[TB ] as
1 − P0
E[TB ] =
KλP0
c−1
1
K ρ c
K!
K
= ρ +n
a n−c
−1 .
Kλ n=0
n c! n=c (K − n)!
17.4 Related works

references: Takacs in 1962 [268], Kleinrock in 1975 [270], Daley and Servi in 1998
[288], Artalejo and Lopez-Herrero in 2001 [289], Elhanany and Kahane in 2005 [290],
Dudin and Kim in 2017 [291], Fidler et al. in 2018 [292], and Jiang et al. in 2018 [293].

[Q17.1] Describe the following multi-server queuing systems: M /M /c,
M /M /c/K, and M /M /c/K/K. Provide an illustrative example for each
of them.
[Q17.2] What is the stability condition for a multi-server queue?
[Q17.3] Derive the steady-state probability that there are n packets in M /M /c,
M /M /c/K, and M /M /c/K/K queues.
[Q17.4] Derive the average number of packets in the system and in the queue of
M /M /c, M /M /c/K, and M /M /c/K/K queues.
[Q17.5] Derive the average response time and waiting time in M /M /c,
M /M /c/K, and M /M /c/K/K queues.
[Q17.6] Derive server utilization and system utilization of M /M /c, M /M /c/K,

and M /M /c/K/K queues.
[Q17.7] Derive the average busy period of the server in M /M /c, M /M /c/K and
M /M /c/K/K queues.
[Q17.8] When is a multi-server queue blocked?
[Q17.9] Derive the blocking probability of an M /M /c/K queue.
[Q17.10] Derive the average number of packets in the source and the utilization of
the source in an M /M /c/K/K queue.
17.6 Problems
[P17.1] Derive the waiting probability of data packets in M /M /c, M /M /c/K, and
M /M /c/K/K queuing systems.
[P17.2] At a company, two mail servers are working in parallel. E-mails arrive on
an average of every 1 min throughout the day. It takes an average of 2 s to
process an e-mail at the servers. Both interarrival time and service time are
exponentially distributed. The queue is assumed to be infinitely large and
the serving discipline is FIFO.
● Find the probability that there is no e-mail to be processed at the servers.
● Find the average number of e-mails waiting in the queue and in the
system.
● Find the average response time of the servers, the average waiting time
of the e-mails, and the average busy time of the servers.
● Find the server and system utilizations.
● Discuss the impacts on performance measures of the queue when a
third e-mail server working at the same speed as the first two servers
is added to the system.
[P17.3] A queuing system consists of three servers with a finite capacity of six
packets. Both the interarrival and service time are exponentially distributed
with the average interarrival of 6 s and the average service time of 9 s.
● Find the probability that all the servers are idle, the blocking probability
of the queuing system, and the waiting probability of the packets.
● Find the average number of packets in the system and in the queue.
● Find the server utilization, system utilization, and the average busy
period of the servers.
[P17.4] Considering a finite source in a four-server queuing system with a limited
capacity of eight packets and the same population size of eight packets.
Both the packet arrival and service follow Poisson point process with an
arrival rate of 20 packets/s and a service rate at each server of 10 packets/s.
● Find the probability that all the servers are busy and the waiting
probability of the packets.
● Find the average number of packets in the system, in the queue, and in
the source.
● Find the server utilization, source utilization, and system utilization.
● Find the average idle period and busy period of the server.
References
[1] D. L. Spohn, T. Brown, and S. Grau, Data Network Design, 3rd ed. New York,
NY, USA: McGraw-Hill, Inc., 2002.
[2] W. Stallings, Data and Computer Communications, 8th ed. Upper Saddle
River, NJ, USA: Prentice-Hall, Inc., 2006.
[3] M. Dye, R. McDonald, and A. Rufi, Network Fundamentals, CCNA
Exploration Companion Guide, 1st ed. Cisco Press, Indiana, USA, 2011.
[4] A. Fraser, “Towards a universal data transport system,” IEEE J. Sel. Areas
Commun., vol. 1, no. 5, pp. 803–816, 1983.
[5] T.-S. Chu and M. Gans, “High speed infrared local wireless communication,”
IEEE Commun. Mag., vol. 25, no. 8, pp. 4–10, 1987.
[6] K. Pahlavan, T. H. Probert, and M. E. Chase, “Trends in local wireless
networks,” IEEE Commun. Mag., vol. 33, no. 3, pp. 88–95, 1995.
[7] B. P. Crow, I. Widjaja, J. G. Kim, and P. T. Sakai, “IEEE 802.11 wireless local
area networks,” IEEE Commun. Mag., vol. 35, no. 9, pp. 116–126, 1997.
[8] J. Woodcock, Step Up to Networking (IT Professional). Microsoft Press, USA,
1999.
[9] K. Clark and K. Hamilton, CCIE Professional Development: Cisco LAN
Switching. Cisco Press, Indiana, USA, 1999.
[10] J. Bray and C. Sturman, Bluetooth: Connect Without Cables, 1st ed. Upper
Saddle River, NJ, USA: Prentice Hall PTR, 2000.
[11] W. Stallings, Wireless Communications and Networks, 1st ed. Prentice Hall
Professional Technical Reference, New Jersey, USA, 2001.
[12] P. Roshan and J. Leary, 802.11 Wireless LAN Fundamentals. Cisco Press,
Indiana, USA, 2003.
[13] M. Decina and E. Scace, “CCITT recommendations on the ISDN: A review,”
IEEE J. Sel. Areas Commun., vol. 4, no. 3, pp. 320–325, 1986.
[14] J. W. Lechleider, “High bit rate digital subscriber lines: A review of HDSL
progress,” IEEE J. Sel. Areas Commun., vol. 9, no. 6, pp. 769–784, 1991.
[15] N. M. Thorpe and D. Ross, X.25 Made Easy. Upper Saddle River, NJ, USA:
Prentice-Hall, Inc., 1992.
[16] D. E. McDysan and D. L. Spohn, ATM: Theory and Application. New York,
NY, USA: McGraw-Hill, Inc., 1994.
[17] C. Brown and A. Malis, “RFC 2427: Multiprotocol Interconnect over Frame
Relay,” IETF, Tech. Rep., 1998.
[18] D. Ginsburg, ATM: Solutions for Enterprise Internetworking, 2nd ed. Boston,
MA, USA: Addison-Wesley Longman Publishing Co., Inc., 1999.
[19] J. C. Bellamy, Digital Telephony, 3rd ed. Wiley, New York, USA, 2000.
[20] ITU-T, “Information technology—open systems interconnection—basic

reference model: The basic model,” ITU, Recommendation X.200, 1994.
[21] V. G. Cerf and R. E. Khan, “A protocol for packet network intercommunica-
tion,” IEEE Trans. Commun., vol. 22, pp. 637–648, 1974.
[22] A. Rodriguez, J. Gatell, J. Karas, and R. Peschke, TCP/IP Tutorial and Techni-
cal Overview, 7th ed. Upper Saddle River, NJ, USA: Prentice-Hall, Inc., 2001.
[23] D. Boggs, J. Shoch, E. Taft, and R. Metcalfe, “Pup: An internetwork
architecture,” IEEE Trans. Commun., vol. 28, no. 4, pp. 612–624, 1980.
[24] D. Teare, Ed., Designing Cisco Networks. Cisco Press, Indiana, USA, 1999.
[25] W. Stallings, Computer Networking with Internet Protocols and Technology.
Pearson, New Jersey, USA, 2003.
[26] M. A. Sportack and J. Fairweather, IP Routing Fundamentals. Cisco Press,
Indiana, USA, 1999.
[27] J. Doyle and J. Carroll, Routing TCP/IP, Volume 1, 2nd ed. Cisco Press,
Indiana, USA, 2005.
[28] R. Graziani and A. Johnson, Routing Protocols and Concepts, CCNA
Exploration Companion Guide. Cisco Press, Indiana, USA, 2007.
[29] L. R. J. Ford, Network Flow Theory. Santa Monica, CA, USA: RAND
Corporation, 1956.
[30] L. R. J. Ford and D. R. Fulkerson, “Maximal flow through a network,” Can.
J. Math., vol. 8, pp. 399–404, 1956.
[31] R. Bellman, “On a routing problem,” Q. Appl. Math., vol. 16, no. 1,
pp. 87–90, 1958.
[32] C. Hedrick, “RFC 1058: Routing Information Protocol,” IETF, Tech. Rep.,
1988.
[33] G. Malkin and R. Minnear, “RFC 2080: RIPng for IPv6,” IETF, Tech. Rep.,
1997.
[34] G. Malkin, “RFC 2453: RIP Version 2,” IETF, Tech. Rep., 1998.
[35] C. L. H. Rutgers, “An Introduction to IGRP (White Paper),” Cisco Systems,
Inc., Tech. Rep. 26825, 1991.
[36] Cisco, “IGRP Metric (Troubleshooting TechNotes),” Cisco Systems, Inc.,
Tech. Rep. 13678, 2005.
[37] T. Clausen and P. Jacquet, “RFC 3626: Optimized Link State Routing
Protocol (OLSR),” IETF, Tech. Rep., 2003.
[38] J. T. Moy, OSPF: Anatomy of an Internet Routing Protocol. Boston, MA,
USA: Addison-Wesley Longman Publishing Co., Inc., 1998.
[39] J. Moy, “RFC 2328: OSPF Version 2,” IETF, Tech. Rep., 1998.
[40] R. Coltun, D. Ferguson, J. Moy, and A. Lindem, “RFC 5340: OSPF for
IPv6,” IETF, Tech. Rep., 2008.
[41] E. W. Dijkstra, “A note on two problems in connexion with graphs,” Numer.
Math., vol. 1, no. 1, pp. 269–271, 1959.
[42] J. J. Garcia-Lunes-Aceves, “Loop-free routing using diffusing computations,”
IEEE/ACM Trans. Networking, vol. 1, no. 1, pp. 130–141, 1993.
[43] I. Pepelnjak, EIGRP Network Design Solutions. Cisco Press, Indiana, USA,
1999.
References 301
[44] D. Savage, J. Ng, S. Moore, D. Slice, P. Paluch, and R. White, “RFC 7868:
Cisco’s Enhanced Interior Gateway Routing Protocol (EIGRP),” IETF, Tech.
Rep., 2016.
[45] Y. Rekhter and P. Gross, “RFC 1772: Application of the Border Gateway
Protocol in the Internet,” IETF, Tech. Rep., 1995.
[46] Y. Rekhter, T. Li, and S. Hares, “RFC 4271: A Border Gateway Protocol 4
(BGP-4),” IETF, Tech. Rep., 2006.
[47] S. Halabi and D. McPherson, Internet Routing Architectures, 2nd ed. Cisco
Press, Indiana, USA, 2000.
[48] R. White, D. McPherson, and S. Sangli, Practical BGP. Redwood City, CA,
USA: Addison Wesley Longman Publishing Co., Inc., 2004.
[49] T. Bates, E. Chen, and R. Chandra, “RFC 4456: BGP Route Reflection: An
Alternative to Full Mesh Internal BGP (iBGP),” IETF, Tech. Rep., 2006.
[50] P. Traina, D. McPherson, and J. Scudder, “RFC 5065: Autonomous System
Confederations for BGP,” IETF, Tech. Rep., 2007.
[51] J. Postel, “RFC 791: Internet Protocol—DARPA Internet Program Protocol
Specification,” IETF, Tech. Rep., 1981.
[52] P. Mockapetris, “RFC 1035: Domain Names—Implementation and
[53] R. Braden, “RFC 1122: Requirements for Internet Hosts—Communication
Layers,” IETF, Tech. Rep., 1989.
[54] K. Egevang and P. Francis, “RFC 1631: The IP Network Address Translator
(NAT),” IETF, Tech. Rep., 1994.
[55] Y. Rekhter, B. Moskowitz, D. Karrenberg, G. J. de Groot, and E. Lear, “RFC
1918: Address Allocation for Private Internets,” IETF, Tech. Rep., 1996.
[56] M. Cotton, L. Vegoda, R. Bonica, and B. Haberman, “RFC 6890: Special-
Purpose IP Address Registries,” IETF, Tech. Rep., 2013.
[57] C. Perkins, “RFC 2003: IP Encapsulation within IP,” IETF, Tech. Rep., 1996.
[58] C. Perkins, “RFC 5944: IP Mobility Support for IPv4, Revised,” IETF, Tech.
Rep., 2010.
[59] J. Touch, “RFC 6864: Updated Specification of the IPv4 ID Field,” IETF,
Tech. Rep., 2013.
[60] J. Mogul, “RFC 917: Internet Subnets,” IETF, Tech. Rep., 1984.
[61] P. Tsuchiya, “RFC 1219: On the Assignment of Subnet Numbers,” IETF,
Tech. Rep., 1991.
[62] T. Pummill and B. Manning, “RFC 1878: Variable Length Subnet Table For
IPv4,” IETF, Tech. Rep., 1995.
[63] Y. Rekhter and T. Li, “RFC 1518: An Architecture for IP Address Allocation
with CIDR,” IETF, Tech. Rep., 1993.
[64] Y. Rekhter, “RFC 1817: CIDR and Classful Routing,” IETF, Tech. Rep.,
1995.
[65] V. Fuller and T. Li, “RFC 4632: Classless Inter-domain Routing (CIDR): The
Internet Address Assignment and Aggregation Plan,” IETF, Tech. Rep., 2006.
[66] R. Hinden and B. Haberman, “RFC 4193: Unique Local IPv6 Unicast
Addresses,” IETF, Tech. Rep., 2005.
[67] R. Hinden and S. Deering, “RFC 4291: IP Version 6AddressingArchitecture,”

IETF, Tech. Rep., 2006.
[68] M. Boucadair and S. Venaas, “RFC 7371: Updates to the IPv6 Multicast
Addressing Architecture,” IETF, Tech. Rep., 2014.
[69] S. Deering and R. Hinden, “RFC 8200: Internet Protocol, Version 6 (IPv6)
[70] D. Minoli, Broadband Network Analysis and Design. Norwood, MA, USA:
Artech House, Inc., 1993.
[71] C. Partridge, Gigabit Networking. Boston, MA, USA: Addison-Wesley
Longman Publishing Co., Inc., 1994.
[72] M. Pióro and D. Medhi, Routing, Flow, and Capacity Design in Communica-
tion and Computer Networks. San Francisco, CA, USA: Morgan Kaufmann
Publishers Inc., 2004.
[73] J. McCabe, Ed., Network Analysis, Architecture, and Design, 3rd ed. Morgan
Kaufmann, Massachusetts, USA, 2007.
[74] K. Raza and M. Turner, Large-scale IP Network Solutions. Cisco Press,
Indiana, USA, 1999.
[75] E. Ravasz and A.-L. Barabasi, “Hierarchical organization in complex
networks,” Phys. Rev. E, vol. 67, no. 2, p. 026112, 2003.
[76] Cisco Networking Academy, Connecting Networks Companion Guide. Cisco
[77] P. Oppenheimer, Top-Down Network Design, 3rd ed. Cisco Press, Indiana,
USA, 2010.
[78] R. White, A. Retana, and D. Slice, Optimal Routing Design, 1st ed. Cisco
[79] D. E. Robling Denning, Cryptography and Data Security. Boston, MA,
USA: Addison-Wesley Longman Publishing Co., Inc., 1982.
[80] W. Stallings, Cryptography and Network Security: Principles and Practice,
3rd ed. Pearson Education, New Jersey, USA, 2002.
[81] S. Kent and K. Seo, “RFC 4301: Security Architecture for the Internet
Protocol,” IETF, Tech. Rep., 2005.
[82] W. Jiang, H. Song, and Y. Dai, “Real-time intrusion detection for high-speed
networks,” Comput. Secur., vol. 24, no. 4, pp. 287–294, 2005.
[83] Y. S. Shiu, S. Y. Chang, H. C. Wu, S. C. H. Huang, and H. H. Chen, “Physical
layer security in wireless networks: A tutorial,” IEEE Wireless Commun.,
vol. 18, no. 2, pp. 66–74, 2011.
[84] H.-J. Liao, C.-H. R. Lin, Y.-C. Lin, and K.-Y. Tung, “Intrusion detection
system: A comprehensive review,” J. Network Comput. Appl., vol. 36, no. 1,
pp. 16–24, 2013.
[85] J. Pescatore, “DDoS attacks advancing and enduring: A SANS survey,” Feb.
2014. [Online]. Available: https://www.sans.org/reading-room/whitepapers/
analyst/ddos-attacks-advancing-enduring-survey-34700.
[86] R. K. Deka, K. P. Kalita, D. Bhattacharya, and J. K. Kalita, “Network
defense: Approaches, methods and techniques,” J. Network Comput. Appl.,
vol. 57, pp. 71–84, 2015.
References 303
[87] T. A. Le, H. X. Nguyen, Q.-T. Vien, and M. Karamanoglu, “Secure informa-

tion transmission and power transfer in cellular networks,” IEEE Commun.
Lett., vol. 19, no. 9, pp. 1532–1535, 2015.
[88] T. A. Le, H. X. Nguyen, Q. T. Vien, and M. Karamanoglu, “Secure infor-
mation transmission in the presence of energy-harvesting eavesdroppers in
multi-cell networks,” in 2015 IEEE Global Communications Conference
(GLOBECOM), 2015, pp. 1–6.
[89] T. A. Le, Q. T. Vien, H. X. Nguyen, D. W. K. Ng, and R. Schober, “Robust
optimization with probabilistic constraints for power-efficient and secure
SWIPT,” in 2016 IEEE Global Communications Conference (GLOBECOM),
2016, pp. 1–7.
[90] T. A. Le, Q.-T. Vien, H. X. Nguyen, D. W. K. Ng, and R. Schober, “Robust
chance-constrained optimization for power-efficient and secure SWIPT
systems,” IEEE Trans. Green Commun. Networking, vol. 1, no. 3, pp.
333–346, 2017.
[91] Q. T. Vien, T. A. Le, H. X. Nguyen, and H. Phan, “A secure network coding
based modify-and-forward scheme for cooperative wireless relay networks,”
in 2016 IEEE 83rd Vehicular Technology Conference (VTC Spring), 2016,
pp. 1–5.
[92] Q. T. Vien, T. A. Le, and T. Q. Duong, “Opportunistic secure transmission for
wireless relay networks with modify-and-forward protocol,” in 2017 IEEE
International Conference on Communications (ICC), 2017, pp. 1–6.
[93] Q. T. Vien, T. A. Le, X. S. Yang, and T. Q. Duong, “On the handover security
key update and residence management in LTE networks,” in 2017 IEEE
Wireless Communications and Networking Conference (WCNC), 2017,
pp. 1–6.
[94] I. Karim and Q.-T. Vien, Snort-based Intrusion Detection System for
Practical Computer Networks: Implementation and Comparative Study.
LAP LAMBERT Academic Publishing, Saarland, Germany, 2017.
[95] D. Bertsekas and R. Gallager, Data Networks, 2nd ed. Upper Saddle River,
NJ, USA: Prentice-Hall, Inc., 1992.
[96] C. White, Data Communications and Computer Networks: A Business User’s
Approach, 6th ed. Boston, MA, USA: Course Technology Press, 2010.
[97] R. Braden, D. Clark, and S. Shenker, “RFC 1633: Integrated Services in the
Internet Architecture: An Overview,” IETF, Tech. Rep., 1994.
[98] G. Huston, “RFC 2990: Next Steps for the IP QoS Architecture,” IETF, Tech.
Rep., 2000.
[99] J. W. Evans and C. Filsfils, Deploying IP and MPLS QoS for Multiservice
Networks: Theory & Practice. San Francisco, CA, USA: Morgan Kaufmann
Publishers Inc., 2007.
[100] M. Marchese, QoS Over Heterogeneous Networks. Wiley Publishing, New
York, USA, 2007.
[101] T. Braun, M. Diaz, J. Enrquez-Gabeiras, and T. Staub, End-to-End Quality
of Service Over Heterogeneous Networks, 1st ed. Springer Publishing
Company, Incorporated, New York, USA, 2008.
[102] D. Papadimitriou, “RFC 6003: Ethernet Traffic Parameters,” IETF, Tech.

Rep., 2010.
[103] F. Baker, J. Polk, and M. Dolly, “RFC 5865: A Differentiated Services Code
Point (DSCP) for Capacity-Admitted Traffic,” IETF, Tech. Rep., 2010.
[104] G. Rubino, Network Reliability Evaluation. Newark, NJ, USA: Gordon and
Breach Science Publishers, Inc., 1999, pp. 275–302.
[105] J. G. Proakis and M. Salehi, Communication Systems Engineering, 2nd ed.
Upper Saddle River, NJ, USA: Prentice-Hall, Inc., 2001.
[106] J. G. Proakis and M. Salehi, Digital Communications, 5th ed. McGraw-Hill,
New York, USA, 2007.
[107] H. Nyquist, “Certain topics in telegraph transmission theory,” Trans. Am.
Inst. Electr. Eng., vol. 47, no. 2, pp. 617–644, 1928.
[108] C. E. Shannon, “A mathematical theory of communication,” Bell Syst. Tech.
J., vol. 27, no. 3, pp. 379–423, 1948.
[109] C. E. Shannon, “Communication in the presence of noise,” Proc. IRE,
vol. 37, no. 1, pp. 10–21, 1949.
[110] R. J. Marks, II, Introduction to Shannon Sampling and Interpolation Theory.
Berlin, Heidelberg: Springer-Verlag, 1991.
[111] T. M. Cover and J. A. Thomas, Elements of Information Theory (Wiley
Series in Telecommunications and Signal Processing). New York, NY, USA:
Wiley-Interscience, 2006.
[112] R. Ramaswamy, N. Weng, and T. Wolf, “Characterizing network processing
delay,” in Global Telecommunications Conference, 2004. GLOBECOM ’04.
IEEE, vol. 3, 2004, pp. 1629–1634.
[113] J. F. Kurose and K. W. Ross, Computer Networking: A Top-Down Approach,
6th ed. Pearson, New Jersey, USA, 2012.
[114] R. Jain, The Art of Computer Systems Performance Analysis: Techniques for
Experimental Design, Measurement, Simulation, and Modeling, 1st ed. John
Wiley & Sons, New York, USA, 1991.
[115] G. Kotonya and I. Sommerville, Requirements Engineering: Processes and
Techniques, 1st ed. Wiley Publishing, New York, USA, 1998.
[116] D. C. Hay, Requirements Analysis: From Business Views to Architecture,
1st ed. Prentice Hall, New Jersey, USA, 2003.
[117] P. A. Laplante, Requirements Engineering for Software and Systems, 2nd ed.
Boston, MA, USA: Auerbach Publications, 2013.
[118] D. Smith, Reliability, Maintainability and Risk, 8th ed.
Butterworth-Heinemann, Oxford, UK, 2011.
[119] M. L. Ayers, Telecommunications System Reliability Engineering, Theory,
and Practice, 1st ed. Wiley-IEEE Press, New York, USA, 2012.
[120] E. A. Elsayed, Reliability Engineering, 2nd ed. Wiley, New York, USA, 2012.
[121] J. Lienig and H. Bruemmer, Reliability Analysis. Cham: Springer
International Publishing, 2017, pp. 45–73.
[122] C. G. Omidyar and G. Pujolle, “Guest editorial—Introduction to flow and
congestion control,” IEEE Commun. Mag., vol. 34, no. 11, pp. 30, 1996.
References 305
[123] N. Brownlee, C. Mills, and G. Ruth, “RFC 2722: Traffic Flow Measurement:
Architecture,” IETF, Tech. Rep., 1999.
[124] S. Floyd and K. Fall, “Promoting the use of end-to-end congestion control
in the Internet,” IEEE/ACM Trans. Networking, vol. 7, no. 4, pp. 458–472,
1999.
[125] A. E. Gamal and Y.-H. Kim, Network Information Theory. New York, NY,
USA: Cambridge University Press, 2012.
[126] P. L. Frana, “Telematics and the early history of international digital
information flows,” IEEE Ann. Hist. Comput., vol. 40, no. 2, pp. 32–47, 2018.
[127] X. Shen, H. Yu, J. Buford, and M. Akon, Handbook of Peer-to-Peer
Networking, 1st ed. Springer Publishing Company, Incorporated, New York,
USA, 2009.
[128] F. H. P. Fitzek and H. Charaf, Mobile Peer-to-Peer Networks: An Introduction
to the Tutorial Guide. Wiley Telecom, New York, USA, 2009.
[129] C. G. Gurler and M. Tekalp, “Peer-to-peer system design for adaptive 3D
video streaming,” IEEE Commun. Mag., vol. 51, no. 5, pp. 108–114, 2013.
[130] N. Kumar and J. H. Lee, “Peer-to-peer cooperative caching for data dissem-
ination in urban vehicular communications,” IEEE Syst. J., vol. 8, no. 4,
pp. 1136–1144, 2014.
[131] Z. Yin, C. Wu, Z. Yang, and Y. Liu, “Peer-to-peer indoor navigation using
smartphones,” IEEE J. Sel. Areas Commun., vol. 35, no. 5, pp. 1141–1153,
2017.
[132] W. Wu, H. Mao, Y. Wang, J. Wang, W. Wang, and C. Tian, “CoolConfer-
encing: Enabling robust peer-to-peer multi-party video conferencing,” IEEE
Access, vol. 5, pp. 25474–25486, 2017.
[133] C. Jiang, L. Gao, L. Duan, and J. Huang, “Scalable mobile crowdsensing
via peer-to-peer data sharing,” IEEE Trans. Mob. Comput., vol. 17, no. 4,
pp. 898–912, 2018.
[134] A. Asadi, Q. Wang, and V. Mancuso, “A survey on device-to-device commu-
nication in cellular networks,” IEEE Commun. Surv. Tutorials, vol. 16, no. 4,
pp. 1801–1819, Fourthquarter 2014.
[135] R. Trestian, Q. T. Vien, H. X. Nguyen, and O. Gemikonakli, “ECO-M:
Energy-efficient cluster-oriented multimedia delivery in a LTE D2D
environment,” in 2015 IEEE International Conference on Communications
(ICC), 2015, pp. 55–61.
[136] Y. Wu, S. Wang, W. Liu, W. Guo, and X. Chu, “Iunius: A cross-layer
peer-to-peer system with device-to-device communications,” IEEE Trans.
Wireless Commun., vol. 15, no. 10, pp. 7005–7017, 2016.
[137] K. Ali, H. X. Nguyen, P. Shah, Q. T. Vien, and N. Bhuvanasundaram,
“Architecture for public safety network using D2D communication,” in 2016
IEEE Wireless Communications and Networking Conference, 2016, pp. 1–6.
[138] K. Ali, H. X. Nguyen, P. Shah, Q. T. Vien, and E. Ever, “D2D multi-hop
relaying services towards disaster communication system,” in 2017 24th
International Conference on Telecommunications (ICT), 2017, pp. 1–5.
[139] K. Ali, H. X. Nguyen, Q. T. Vien, P. Shah, and Z. Chu, “Disaster management

using D2D communication with power transfer and clustering techniques,”
IEEE Access, vol. 6, pp. 14643–14654, 2018.
[140] A. Sinha, “Client-server computing,” Commun. ACM, vol. 35, no. 7,
pp. 77–98, 1992.
[141] S. Ramesh and H. G. Perros, “A multilayer client-server queueing network
model with synchronous and asynchronous messages,” IEEE Trans. Softw.
Eng., vol. 26, no. 11, pp. 1086–1100, 2000.
[142] V. Kanitkar and A. Delis, “Real-time processing in client-server databases,”
IEEE Trans. Comput., vol. 51, no. 3, pp. 269–288, 2002.
[143] I. Abdul-Fatah and S. Majumdar, “Performance of CORBA-based client-
server architectures,” IEEE Trans. Parallel Distrib. Syst., vol. 13, no. 2,
pp. 111–127, 2002.
[144] Y. Zhu, W. Wu, and D. Li, “Efficient client assignment for client-server
systems,” IEEE Trans. Netw. Serv. Manage., vol. 13, no. 4, pp. 835–847,
2016.
[145] J. M. Taylor, “Co-operative computing and control,” IEE Proc. E— Comput.
Digital Tech., vol. 137, no. 1, pp. 1–16, 1990.
[146] T. Richardson, Q. Stafford-Fraser, K. R. Wood, and A. Hopper, “Virtual
network computing,” IEEE Internet Comput., vol. 2, no. 1, pp. 33–38,
1998.
[147] Q. Wu, B. Qin, L. Zhang, J. Domingo-Ferrer, and J. A. Manjon, “Fast trans-
mission to remote cooperative groups: A new key management paradigm,”
[148] C. You and K. Huang, “Exploiting non-causal CPU-state information for
energy-efficient mobile cooperative computing,” IEEE Trans. Wireless
Commun., vol. 17, no. 6, pp. 4104–4117, 2018.
[149] Z. Sheng, C. Mahapatra, V. C. M. Leung, M. Chen, and P. K. Sahu, “Energy
efficient cooperative computing in mobile wireless sensor networks,” IEEE
Trans. Cloud Comput., vol. 6, no. 1, pp. 114–126, 2018.
[150] T. L. Casavant and J. G. Kuhl, “A taxonomy of scheduling in general-purpose
distributed computing systems,” IEEE Trans. Software Eng., vol. 14, no. 2,
pp. 141–154, 1988.
[151] H. S. Morse, Practical Parallel Computing. San Diego, CA, USA: Academic
Press Professional, Inc., 1994.
[152] G. Bianchi, “Performance analysis of the IEEE 802.11 distributed coordi-
nation function,” IEEE J. Sel. Areas Commun., vol. 18, no. 3, pp. 535–547,
2000.
[153] O. Younis and S. Fahmy, “HEED: A hybrid, energy-efficient, distributed
clustering approach for ad hoc sensor networks,” IEEE Trans. Mob. Comput.,
vol. 3, no. 4, pp. 366–379, 2004.
[154] H. Nishida and T. Nguyen, “Optimal client-server assignment for internet
distributed systems,” IEEE Trans. Parallel Distrib. Syst., vol. 24, no. 3,
pp. 565–575, 2013.
References 307
[155] P. Kandaswamy, J. A. Flint, and V. A. Chouliaras, “System on fabrics

architecture using distributed computing,” IEEE Sens. J., vol. 18, no. 14,
pp. 5929–5936, 2018.
[156] E. D. Pascale, I. Macaluso, A. Nag, M. Kelly, and L. Doyle, “The network
as a computer: A framework for distributed computing over IoT mesh
networks,” IEEE Internet Things J., vol. 5, no. 3, pp. 2107–2119, 2018.
[157] D. Estrin, J. C. Mogul, and G. Tsudik, “Visa protocols for controlling
interorganizational datagram flow,” IEEE J. Sel. Areas Commun., vol. 7,
no. 4, pp. 486–498, 1989.
[158] B. Quoitin, C. Pelsser, L. Swinnen, O. Bonaventure, and S. Uhlig, “Inter-
domain traffic engineering with BGP,” IEEE Commun. Mag., vol. 41, no. 5,
pp. 122–128, 2003.
[159] C. Partridge, “RFC 1363: A Proposed Flow Specification,” IETF, Tech. Rep.,
1992.
[160] P. Marques, N. Sheth, R. Raszuk, B. Greene, J. Mauch, and D. McPherson,
“RFC 5575: Dissemination of Flow Specification Rules,” IETF, Tech. Rep.,
2009.
[161] R. L. Cruz, “A calculus for network delay. ii. Network analysis,” IEEE Trans.
Inf. Theory, vol. 37, no. 1, pp. 132–141, 1991.
[162] A. K. Parekh and R. G. Gallager, “A generalized processor sharing approach
to flow control in integrated services networks: The single-node case,”
[163] C. C. Jane, J. S. Lin, and J. Yuan, “Reliability evaluation of a limited-flow
network in terms of minimal cutsets,” IEEE Trans. Reliab., vol. 42, no. 3,
pp. 354–361, 368, 1993.
[164] S. Sen and J. Wang, “Analyzing peer-to-peer traffic across large networks,”
[165] P. Popovski and H. Yomo, “Wireless network coding by amplify-and-forward
for bi-directional traffic flows,” IEEE Commun. Lett., vol. 11, no. 1,
pp. 16–18, 2007.
[166] S. Katti, H. Rahul, W. Hu, D. Katabi, M. Medard, and J. Crowcroft, “XORs
in the air: Practical wireless network coding,” IEEE/ACM Trans. Networking,
vol. 16, no. 3, pp. 497–510, 2008.
[167] K. Sohn and D. Kim, “Dynamic origin-destination flow estimation using
cellular communication system,” IEEE Trans. Veh. Technol., vol. 57, no. 5,
pp. 2703–2713, 2008.
[168] Q. Liu and C. W. Chen, “Smart downlink scheduling for multimedia
streaming over LTE networks with hard handoff,” IEEE Trans. Circuits Syst.
Video Technol., vol. 25, no. 11, pp. 1815–1829, 2015.
[169] W. J. Dally, “Virtual-channel flow control,” IEEE Trans. Parallel Distrib.
Syst., vol. 3, no. 2, pp. 194–205, 1992.
[170] N. A. Jagadeesan and B. Krishnamachari, “Software-defined networking
paradigms in wireless networks: A survey,” ACM Comput. Surv., vol. 47,
no. 2, pp. 27:1–27:11, 2014.
[171] A. Lara, A. Kolasani, and B. Ramamurthy, “Network innovation using

OpenFlow: A survey,” IEEE Commun. Surv. Tutorials, vol. 16, no. 1, pp.
493–512, First 2014.
[172] D. Kreutz, F. M. V. Ramos, P. E. Verissimo, C. E. Rothenberg, S. Azodol-
molky, and S. Uhlig, “Software-defined networking: A comprehensive
survey,” Proc. IEEE, vol. 103, no. 1, pp. 14–76, 2015.
[173] M. A. Hassan, Q.-T. Vien, and M. Aiash, “Software defined networking
for wireless sensor networks: A survey,” Adv. Wireless Commun. Networks,
vol. 3, no. 2, pp. 10–22, 2017.
[174] G. Li, H. Zhou, B. Feng, and G. Li, “Context-aware service function chaining
and its cost-effective orchestration in multi-domain networks,” IEEE Access,
vol. 6, pp. 34976–34991, 2018.
[175] Y. Jia, C. Wu, Z. Li, F. Le, and A. Liu, “Online scaling of NFV service chains
across Geo-distributed datacenters,” IEEE/ACM Trans. Networking, vol. 26,
no. 2, pp. 699–710, 2018.
[176] J. Gray, Benchmark Handbook: For Database and Transaction Processing
Systems. San Francisco, CA, USA: Morgan Kaufmann Publishers Inc.,
1992.
[177] R. Madhavan, E. Tunstel, and E. Messina, Performance Evaluation and
Benchmarking of Intelligent Systems, 1st ed. Springer Publishing Company,
Incorporated, New York, USA, 2009.
[178] M. Harchol-Balter, Performance Modeling and Design of Computer Sys-
tems: Queueing Theory in Action, 1st ed. New York, NY, USA: Cambridge
University Press, 2013.
[179] R. Nambiar and M. Poess, Performance Evaluation and Benchmarking: Tra-
ditional to Big Data to Internet of Things. Springer International Publishing,
New York, USA, 2016.
[180] Wolfram, “Mathematica.” [Online]. Available: http://www.wolfram.com/
mathematica/.
[181] MathWorks, “Matlab.” [Online]. Available: https://uk.mathworks.com/
products/matlab.html.
[182] M. L. Abell and J. P. Braselton, Mathematica by Example, 5th ed. Orlando,
FL, USA: Academic Press, Inc., 2017.
[183] S. Wolfram, An Elementary Introduction to the Wolfram Language, 2nd ed.
Wolfram Media, Inc., 2017.
[184] A. Varga, “Using the OMNeT++ discrete event simulation system in
education,” IEEE Trans. Educ., vol. 42, no. 4, 11 pp, 1999.
[185] J. Banks, J. S. Carson, B. L. Nelson, and D. M. Nicol, Discrete-Event System
Simulation, 5th ed. Pearson, New Jersey, USA, 2010.
[186] K. Wehrle, M. Gnes, and J. Gross, Modeling and Tools for Network Simula-
tion, 1st ed. Springer Publishing Company, Incorporated, New York, USA,
2010.
[187] T. A. Davis, MATLAB Primer, 8th ed. Boca Raton, FL, USA: CRC Press,
Inc., 2010.
[188] B. Rimoldi, Principles of Digital Communication: A Top-Down Approach.
Cambridge University Press, Cambridge, UK, 2016.
References 309
[189] K. Choi and H. Liu, Problem-Based Learning in Communication Systems

Using MATLAB and Simulink. John Wiley & Sons, Inc., New York, USA,
2016.
[190] S. Attaway, MATLAB: A Practical Introduction to Programming and Problem
Solving, 4th ed. Elsevier Science, Amsterdam, Netherlands, 2017.
[191] NS-3 Consortium, “ns-3.” [Online]. Available: https://www.nsnam.org.
[192] OpenSim Ltd., “Omnet++.” [Online]. Available: https://omnetpp.org.
[193] MathWorks, “Simulink.” [Online]. Available: https://uk.mathworks.com/
products/simulink.html.
[194] J. Walrand, K. Bagchi, and G. W. Zobrist, Eds., Network Performance
Modeling and Simulation. Newark, NJ, USA: Gordon and Breach Science
Publishers, Inc., 1999.
[195] S. Robinson, Simulation: The Practice of Model Development and Use.
USA: John Wiley & Sons, Inc., New York, USA, 2004.
[196] D. Murray-Smith, Modelling and Simulation of Integrated Systems in
Engineering: Issues of Methodology, Quality, Testing and Application.
Woodhead Publishing, Limited, Cambridge, UK, 2012.
[197] B. Kumar and E. S. Davidson, “Performance evaluation of highly concurrent
computers by deterministic simulation,” Commun. ACM, vol. 21, no. 11,
pp. 904–913, 1978.
[198] D. Poole and A. E. Raftery, “Inference for deterministic simulation models:
The Bayesian melding approach,” J. Am. Stat. Assoc., vol. 95, no. 452,
pp. 1244–1255, 2000.
[199] J. P. C. Kleijnen, Design and Analysis of Simulation Experiments, 1st ed.
Springer Publishing Company, Incorporated, New York, USA, 2007.
[200] H. C. Tijms, Stochastic Modelling and Analysis: A Computational Approach.
New York, NY, USA: John Wiley & Sons, Inc., 1986.
[201] B. D. Ripley, Stochastic Simulation. New York, NY, USA: John Wiley &
Sons, Inc., 1987.
[202] V. G. Kulkarni, Modeling and Analysis of Stochastic Systems. London, UK:
Chapman & Hall, Ltd., 1995.
[203] M. A. Pinsky and S. Karlin, An Introduction to Stochastic Modeling. Orlando,
FL, USA: Academic Press, Inc., 2011.
[204] G. S. Fishman, Discrete-Event Simulation. Berlin, Heidelberg: Springer-
Verlag, 2001.
[205] B. K. Choi and D. Kang, Modeling and Simulation of Discrete Event Systems,
1st ed. Wiley Publishing, New York, USA, 2013.
[206] B. P. Zeigler, T. G. Kim, and H. Praehofer, Theory of Modeling and
Simulation, 2nd ed. Orlando, FL, USA: Academic Press, Inc., 2000.
[207] F. E. Cellier and E. Kofman, Continuous System Simulation. Berlin,
Heidelberg: Springer-Verlag, 2006.
[208] L. G. Birta and G. Arbez, Modelling and Simulation: Exploring Dynamic
System Behaviour. Berlin, Heidelberg: Springer-Verlag, 2007.
[209] S. Bandyopadhyay and R. Bhattacharya, Discrete and Continuous Simu-
lation: Theory and Practice, 1st ed. Boca Raton, FL, USA: CRC Press,
Inc., 2014.
[210] A. C. Davison, Statistical Models. Cambridge University Press, Cambridge,

UK, 2003.
[211] D. R. Cox, Principles of Statistical Inference. Cambridge University Press,
Cambridge, UK, 2006.
[212] J. Rissanen, Information and Complexity in Statistical Modeling, 1st ed.
[213] L. Wasserman, All of Statistics: A Concise Course in Statistical Inference.
[214] D. P. Kroese and J. C. C. Chan, Statistical Modeling and Computation.
[215] A. Papoulis, Probability, Random Variables, and Stochastic Processes,
4th ed. McGraw Hill, New York, USA, 2002.
[216] C. Forbes, M. Evans, N. Hastings, and B. Peacock, Statistical Distributions,
4th ed. Wiley, New York, USA, 2010.
[217] D. C. Montgomery and G. C. Runger, Applied Statistics and Probability for
Engineers, 6th ed. Wiley, New York, USA, 2013.
[218] S. M. Ross, Introduction to Probability Models, 11th ed. Orlando, FL, USA:
Academic Press, Inc., 2014.
[219] S. M. Ross, Stochastic Processes, 2nd ed. Wiley, New York, USA, 1995.
[220] R. Nelson, Probability, Stochastic Processes, and Queueing Theory: The
Mathematics of Computer Performance Modeling. Berlin, Heidelberg:
Springer-Verlag, 1995.
[221] D. Daley and D. Vere-Jones, An Introduction to the Theory of Point Processes:
Volume I: Elementary Theory and Methods, 2nd ed. New York, NY, USA:
Springer-Verlag New York, 2003.
[222] D. Daley and D. Vere-Jones, An Introduction to the Theory of Point Processes:
Volume II: General Theory and Structure, 2nd ed. New York, NY, USA:
Springer-Verlag New York, 2008.
[223] J. Illian, A. Penttinen, H. Stoyan, and D. Stoyan, Statistical Analysis and Mod-
elling of Spatial Point Patterns. John Wiley & Sons, New York, USA, 2008.
[224] R. F. Bass, Stochastic Processes, 1st ed. New York, NY, USA: Cambridge
University Press, 2011.
[225] D. Snyder and M. Miller, Random Point Processes in Time and Space,
2nd ed. New York, NY, USA: Springer New York, 2012.
[226] M. S. Obaidat, F. Zarai, and P. Nicopolitidis, Modeling and Simulation of
Computer Networks and Systems: Methodologies and Applications, 1st ed.
San Francisco, CA, USA: Morgan Kaufmann Publishers Inc., 2015.
[227] A. van den Bos, Parameter Estimation for Scientists and Engineers.
New York, NY, USA: Wiley-Interscience, 2007.
[228] B. C. Levy, Principles of Signal Detection and Parameter Estimation, 1st ed.
[229] R. C. Aster, B. Borchers, and C. H. Thurber, Parameter Estimation and
Inverse Problems, 2nd ed. Academic Press, Massachusetts, USA, 2012.
[230] F. J. Massey, “The Kolmogorov–Smirnov test for goodness of fit,” J. Am.
Stat. Assoc., vol. 46, no. 253, pp. 68–78, 1951.
References 311
[231] R. B. D’Agostino and M. A. Stephens, Eds., Goodness-of-fit Techniques.

New York, NY, USA: Marcel Dekker, Inc., 1986.
[232] P. E. Greenwood and M. S. Nikulin, Eds., A Guide to Chi-Squared Testing.
Wiley, New York, USA, 1996.
[233] Y. Fan, “Goodness-of-fit tests for a multivariate distribution by the empirical
characteristic function,” J. Multivar. Anal., vol. 62, no. 1, pp. 36–63,
1997.
[234] A. Justel, D. Pena, and R. Zamar, “A multivariate Kolmogorov–Smirnov test
of goodness of fit,” Stat. Probab. Lett., vol. 35, no. 3, pp. 251–259, 1997.
[235] W. Feller, An Introduction to Probability Theory and Its Applications, Volume
1, 3rd ed. Wiley, New York, USA, 1968.
[236] R. D. Yates and D. J. Goodman, Probability and Stochastic Processes: A
Friendly Introduction for Electrical and Computer Engineers, 2nd ed. John
[237] J. A. Gubner, Probability and Random Processes for Electrical and Computer
Engineers. New York, NY, USA: Cambridge University Press, 2006.
[238] A. Leon-Garcia, Probability and Random Processes For EE’s, 3rd ed. Upper
Saddle River, NJ, USA: Prentice-Hall, Inc., 2007.
[239] R. Durrett, Probability: Theory and Examples, 4th ed. New York, NY, USA:
Cambridge University Press, 2010.
[240] R. G. Gallager, Stochastic Processes: Theory for Applications. Cambridge
University Press, Cambridge, UK, 2014.
[241] R. Steyer and W. Nagel, Probability and Conditional Expectation:
Fundamentals for the Empirical Sciences. Wiley, New York, USA, 2017.
[242] M. Mitzenmacher and E. Upfal, Probability and Computing: Randomization
and Probabilistic Techniques in Algorithms and Data Analysis, 2nd ed. New
York, NY, USA: Cambridge University Press, 2017.
[243] Y. V. Prohorov and Y. A. Rozanov, Probability Theory: Basic Concepts, Limit
Theorems, Random Processes, 1st ed. Springer-Verlag Berlin Heidelberg,
Berlin, Germany, 1969.
[244] N. L. Johnson, S. Kotz, and N. Balakrishnan, Continuous Univariate
Distributions, Volume 1, 2nd ed. Wiley, New York, USA, 1994.
[245] N. L. Johnson, A. W. Kemp, and S. Kotz, Univariate Discrete Distributions,
3rd ed. John Wiley & Sons, Inc., New York, USA, 2005.
[246] W. J. Stewart, Probability, Markov Chains, Queues, and Simulation:
The Mathematical Basis of Performance Modeling. Princeton, NJ, USA:
Princeton University Press, 2009.
[247] J. E. Gentle, Computational Statistics, 1st ed. Springer Publishing Company,
New York, USA, Incorporated, 2009.
[248] J. Miller, Table of Binomial Coefficients (Royal Society Mathematical Tables:
Volume 3). Cambridge University Press, Cambridge, UK, 1954.
[249] R. Kaas and J. Buhrman, “Mean, median and mode in binomial distributions,”
Stat. Neerl., vol. 34, no. 1, pp. 13–18, 1980.
[250] Y. H. Wang, “On the number of successes in independent trials,” Stat. Sin.,
vol. 3, no. 2, pp. 295–312, 1993.
[251] G. E. Andrews, “The geometric series in calculus,” Am. Math. Mon., vol. 105,
no. 1, pp. 36–40, 1998.
[252] M. R. Spiegel, Theory and Problems of probability and Statistics (Schaum’s
Outline Series). McGraw-Hill Education (India) Pvt Limited, 2003. [Online].
Available: https://books.google.co.uk/books?id=qXPAOq5qE5wC.
[253] H. S. Wilf, Generatingfunctionology. Natick, MA, USA: A. K. Peters, Ltd.,
2006.
[254] J. Hilbe, Negative Binomial Regression. Cambridge University Press,
Cambridge, UK, 2011.
[255] F. Haight, Handbook of the Poisson distribution, ser. Publications in
Operations Research. Wiley, New York, USA, 1967.
[256] L. Knüsel, “Computation of the Chi square and Poisson distribution,” SIAM
J. Sci. Stat. Comput., vol. 7, no. 3, pp. 1022–1036, 1986.
[257] N. Balakrishnan and A. P. Basu, Exponential Distribution: Theory, Methods
and Applications. Taylor & Francis, Oxford, UK, 1996.
[258] M. V. Jambunathan, “Some properties of beta and gamma distributions,”
Ann. Math. Stat., vol. 25, no. 2, pp. 401–405, 1954.
[259] J. Patel and C. Read, Handbook of the Normal Distribution, 2nd ed., ser.
Statistics: A Series of Textbooks and Monographs. Taylor & Francis, Oxford,
UK, 1996.
[260] J. Aitchison and J. Brown, Lognormal Distribution, ser. Department of
Applied Economics Monographs. Cambridge University Press, Cambridge,
UK, 1957.
[261] E. L. Crow and K. Shimizu, Lognormal Distributions: Theory and Appli-
cations, ser. Statistics: A Series of Textbooks and Monographs. CRC Press,
Florida, USA, 1987.
[262] G. E. P. Box and M. E. Muller, “A note on the generation of random normal
deviates,” Ann. Math. Stat., vol. 29, no. 2, pp. 610–611, 1958.
[263] L. Devroye, “On the computer generation of random variables with a given
characteristic function,” Comput. Math. Appl., vol. 7, no. 6, pp. 547–552,
1981.
[264] L. Devroye, Non-Uniform Random Variate Generation, 1st ed. Springer-
Verlag Berlin, Germany, 1986.
[265] L. Barabesi and L. Pratelli, “Universal methods for generating random
variables with a given characteristic function,” J. Stat. Comput. Simul.,
vol. 85, no. 8, pp. 1679–1691, 2015.
[266] C. P. Robert and G. Casella, Monte Carlo Statistical Methods (Springer Texts
in Statistics). Berlin, Heidelberg: Springer-Verlag, 2005.
[267] R. Y. Rubinstein and D. P. Kroese, Simulation and the Monte Carlo Method,
ser. Wiley Series in Probability and Statistics, 2nd ed. Wiley, New York,
USA, 2007.
[268] L. Takács, Introduction to the Theory of Queues, ser. University Texts in the
Mathematical Sciences. Oxford University Press, Oxford, UK, 1962.
[269] G. Newell, Applications of Queueing Theory, ser. Monographs on Applied
Probability and Statistics. Chapman and Hall, London, UK, 1971.
References 313
[270] L. Kleinrock, Queueing Systems: Volume 1: Theory. New York, NY, USA:
Wiley-Interscience, 1975.
[271] L. Kleinrock, Queueing Systems: Volume 2: Computer Applications. John
[272] G. Giambene, Queuing Theory and Telecommunications: Networks and
Applications. Berlin, Heidelberg: Springer-Verlag, 2005.
[273] A. S. Alfa, Queueing Theory for Telecommunications: Discrete Time
Modelling of a Single Node System, 1st ed. Springer Publishing Company,
Incorporated, New York, USA, 2010.
[274] R. W. Wolff, “Poisson arrivals see time averages,” Oper. Res., vol. 30, no. 2,
pp. 223–231, 1982.
[275] E. A. van Doorn and G. Regterschot, “Conditional pasta,” Oper. Res. Lett.,
vol. 7, no. 5, pp. 229–232, 1988.
[276] J. D. C. Little, “A proof for the queuing formula: L = λw,” Oper. Res., vol. 9,
no. 3, pp. 383–387, 1961.
[277] S. Stidham, “A last word on L = λw,” Oper. Res., vol. 22, no. 2, pp. 417–421,
1974.
[278] D. G. Kendall, “Stochastic processes occurring in the theory of queues and
their analysis by the method of the imbedded Markov chain,” Ann. Math.
Stat., vol. 24, no. 3, pp. 338–354, 1953.
[279] A. Lee, Applied Queueing Theory, ser. Studies in Management. Macmillan,
London, UK, 1966.
[280] L. Takács, “A single-server queue with feedback,” Bell Syst. Tech. J., vol. 42,
no. 2, pp. 505–519, 1963.
[281] W. Chang, “Single-server queuing processes in computing systems,” IBM
Syst. J., vol. 9, no. 1, pp. 36–71, 1970.
[282] B. Gopinath and J. A. Morrison, “Discrete-time single server queues with
correlated inputs,” Bell Syst. Tech. J., vol. 56, no. 9, pp. 1743–1768,
1977.
[283] A. Eckberg, “The single server queue with periodic arrival process and deter-
ministic service times,” IEEE Trans. Commun., vol. 27, no. 3, pp. 556–562,
1979.
[284] H. Wang and N. B. Mandayam, “A simple packet-transmission scheme for
wireless data over fading channels,” IEEE Trans. Commun., vol. 52, no. 7,
pp. 1055–1059, 2004.
[285] A. Steiner and S. Shamai, “On queueing and multilayer coding,” IEEE Trans.
Inf. Theory, vol. 56, no. 5, pp. 2392–2415, 2010.
[286] J. Cohen, The Single Server Queue, 2nd ed., ser. North-Holland Series
in Applied Mathematics and Mechanics. Elsevier Science, Amsterdam,
Netherlands, 2012.
[287] L. Zhao, W. Sun, Y. Shi, and J. Liu, “Optimal placement of cloudlets for
access delay minimization in SDN-based Internet of Things networks,” IEEE
Internet Things J., vol. 5, no. 2, pp. 1334–1344, 2018.
[288] D. J. Daley and L. D. Servi, “Idle and busy periods in stable M/M/k queues,”
J. Appl. Probab., vol. 35, no. 4, pp. 950–962, 1998.
[289] J. R. Artalejo and M. J. Lopez-Herrero, “Analysis of the busy period for the
M/M/c queue: An algorithmic approach,” J. Appl. Probab., vol. 38, no. 1,
pp. 209–222, 2001.
[290] I. Elhanany and M. Kahane, “Heterogeneous bursty traffic dispersion over
multiple server clusters,” IEEE Commun. Lett., vol. 9, no. 3, pp. 261–263,
2005.
[291] S. Dudin and C. Kim, “Analysis of multi-server queue with spatial generation
and location-dependent service rate of customers as a cell operation model,”
IEEE Trans. Commun., vol. 65, no. 10, pp. 4325–4333, 2017.
[292] M. Fidler, B. Walker, and Y. Jiang, “Non-asymptotic delay bounds for
multi-server systems with synchronization constraints,” IEEE Trans. Parallel
Distrib. Syst., vol. 29, no. 7, pp. 1545–1559, 2018.
[293] Z. Jiang, B. Krishnamachari, S. Zhou, and Z. Niu, “Optimal sleeping
mechanism for multiple servers with MMPP-based bursty traffic arrival,”
IEEE Wireless Commun. Lett., vol. 7, no. 3, pp. 436–439, 2018.
Index
access layer 85–6 buffered repeaters 39

adaptive routing 42 bus topology 7
Address Resolution Protocol (ARP) 32,
34 C++ programming language 130
analytical modeling 131–2 campus area network (CAN) 2–3
anycast address 72 capacity 95, 100, 104
application-layer protocols 34–5 capacity planning 115–16
application-requirement analysis capital expenditures (CAPEX) 93
100–1 carrier equipment 26
asymmetric DSL (ADSL) 27 carrier sense multiple access with
asynchronous transfer mode (ATM) collision avoidance (CSMA/CA)
25–6 13
autonomous system (AS) 36 carrier sense multiple access with
availability, defined 104 collision detection (CSMA/CD)
12–13
backbone flow 116, 122, 126 Classes of Service (CoS) 95
bandwidth, defined 5 classful IP addresses 63–6
banking system, components in 134 classful routing protocol 50
Basic Rate Interface (BRI) 24 classless inter-domain routing (CIDR)
Bayes’ theorem 165–7 notation 62
benchmarking 129–30 client–server model 119, 126
Bernoulli random variable 197–9, 202, closely coupled nodes 120
242–3 cloud computing applications 101
Bernoulli trial 198, 200 coarse granularity 120–1
binomial distribution 153 collectively exhaustive set 160
binomial random variable 199–202, Committed Information Rates (CIRs)
248 95
birth–death process composite flow 116
steady-state distribution of 261 computer and communication networks,
see also Poisson point process (PPP), queuing systems in 256–7
for queuing systems computers 39
bit error rate (BER) 95–6 computing cluster 121
Border Gateway Protocols (BGPs) 47, computing nodes 120–1
56–7 conditional pmf 178
bridges/switches 38, 40 conditional probability 164–8
broadcast transmission 15 configuration of network 90
connectionless packet switching 21 design process 82

connection-oriented packet switching deterministic simulation models 139
21 device-to-device (D2D) communication
content-delivery network (CDN) 87 systems 118
continuous distributions 146 digital subscriber line (DSL) 26–8
continuous random variables 186–92, Dijkstra’s algorithm 55–6
217 directly connected networks 47
continuous uniform random variable discrete distributions 146, 153
217–21 discrete random variables 176–86, 197
Erlang random variable 226–9 Bernoulli random variable 197–9,
exponential random variable 222–6 202
formulas of 218 binomial random variable 199–202
lognormal random variable 235–8 discrete uniform random variable
normal random variable 230–4 211–14
review of 217 formulas of 198
continuous simulation models 140 geometric random variable 202–5
continuous uniform distribution 153 Pascal random variable 205–8
continuous uniform random variable Poisson random variable 208–11
187–8, 217, 246 review of 197
expected value and variance of discrete simulation models 139
220–1 discrete uniform random variable
controlled-rate application 100–1 211–14, 245
control token 13–14 distance-vector routing algorithms 47
cooperative computing model 119–20, distance-vector routing protocols
126 (DVRPs) 49
core layer 84 Interior Gateway Routing Protocol
cost versus performance 82 (IGRP) 50–1
count-to-infinity condition 52 Routing Information Protocol (RIP)
count-to-infinity problem 50 50
covariance 184 routing loops 51–2
cumulative distribution function (cdf) distributed computing model 120–1,
178–80, 187–8 126
of continuous uniform random distribution fitness test 155
variable 219–20 distribution layer 84–5
of Erlang random variable 228–9 domain name system (DNS) 33, 35
of exponential random variable dotted-quad format 63
222–4 dynamic routing algorithms 47
of lognormal random variable 237 dynamic routing protocols 48–9
of normal random variable 231–3 dynamic simulation models 139
customer premises equipment (CPE) 26
empirical distribution 153
datagram packet switching 21 encapsulation 31
data sources and sinks 117 Enhanced Interior Gateway Routing
data transport applications 101 Protocol (EIGRP) 45, 56
design approaches and evolution 86–7 enterprise requirements 102
Index 317
Erlang random variable 226–9, 250 Gaussian random variable: see normal
expected value and variance of 229 random variable
ES 36 generic computing devices 101
event, defined 160–1 geometric random variable 202–5,
event notification via monitoring 90 243–4
exchange-termination equipment 23
expected value 180–2 half-duplex systems 117
of continuous random variable 190 hierarchical design of network 49, 82–3
of discrete random variables 180–2 hierarchical layers of network 84
experiment, defined 159 access layer 85–6
exponential distribution 145–6, 153 core layer 84
exponential random variable 222–6, distribution layer 84–5
247 high-bit-rate DSL 28
expected value and variance of hold-down timers 52
224–5 home area network (HAN) 2
exterior gateway protocol (EGP) 47–8 homogeneous Poisson point process
147–50
failure rate 104
host-requirement analysis 101–2
fast Ethernet 11
hosts 93
File Transfer Protocol (FTP) 29, 35
hubs 5–6
financial requirements 102
100BaseTX and 100BaseFX fast
fine granularity 120
Ethernet 11
finest grain set 160
HyperText Transfer Protocol (HTTP)
flat routing algorithms 47
35
flow analysis model 125–6
flow boundaries 121–2
flow distributions 122–3 independence of events 168–70
flow models 117 individual flow 116
client–server model 119 inhomogeneous Poisson point process
cooperative computing model 147
119–20 input modelling 152
distributed computing model 120–1 data collection 152–3
peer-to-peer (P2P) model 118 distribution fitness test 155
flows 115–16 parameter estimation 154–5
flowspec 123–4 probability distribution identification
developing 126 153–4
flow specification 123–5 Integrated Services Digital Network
flow types 116 (ISDN) 22–4
frame relay 24–5 interarrival and service time in Poisson
full-duplex system 117 point process 259
full mesh topologies 19–20 interior gateway protocols (IGPs) 47,
49–56
Gamma and Erlang distribution 146 Interior Gateway Routing Protocol
Gamma and Weibull distribution 146 (IGRP) 45, 50–1
gateways 40 intermediate system (IS) 36
Intermediate System to Intermediate information exchange and

System (IS–IS) 53 information formats in OSI
International Organization for layers 31–2
Standardization (ISO) 28 interaction between OSI layers in
Internet Control Message Protocol communications 30–1
(ICMP) 32, 34 OSI layers 29–30
Internet Engineering Task Force (IETF) wide area networks (WAN) 16
94 devices 16–18
Internet Protocol addressing 37, 61 technologies 20–8
addressing mechanisms 63 topologies 18–20
addressing types 63 inverse transform technique 241
IP version 4 (IPv4) 63–4 Bernoulli random variable generation
classful IP addresses 64–6 242–3
private IP addresses 64 continuous uniform random variable
IP version 6 71–3 generation 246
subnetting 66–8 discrete uniform random variable
supernetting 70–1 generation 245
variable-length subnetting (VLSM) exponential random variable
69–70 generation 247
Internet Protocols 32 geometric random variable
application-layer protocols 35 generation 243–4
ARP 34 IP version 4 (IPv4) 63–4
ICMP 34 classful IP addresses 64–6
Internet Protocol 34 private IP addresses 64
TCP/IP suite 33 IP version 6 71–3
Transmission Control Protocol (TCP)
34 Kendall’s notation for queuing systems
User datagram protocol (UDP) 34 262–3
Internet service providers (ISPs) 56 Kolmogorov–Smirnov (K–S) test 155
internetworking and network kurtosis 186
fundamentals 1–5
Internet protocols 32–5 latency 95–6, 100, 104
internetworks 35 layer-distributing network services 84
addressing 36–7 line-termination equipment 23
devices 38–40 link-state advertisements (LSAs) 53
hierarchy 36 link-state routing protocols (LSRPs)
standards organizations 38 49, 52
local area networks (LAN) 5 Dijkstra’s algorithm 55–6
access methods 12–14 Open Shortest Path First (OSPF)
devices 5–7 53–4
technologies 9–12 Little’s law 259
topologies 7–8 for queuing systems 261–2
transmission methods 14–16 local addresses versus global addresses
open systems interconnection (OSI) 63
reference model 28–9 local area network (LAN) 1–2
Index 319
access methods 12 network-management design 89–91

carrier sense multiple access with technology evaluation criteria 89
collision avoidance (CSMA/CA) management requirements 102
13 Markov process 147
carrier sense multiple access with MATHEMATICA 131
collision detection (CSMA/CD) mathematical modelling: see analytical
12–13 modeling
control token 13–14 MATLAB 131
devices 5 maximum bit rate 95
hubs 5–6 mean time between failures (MTBF)
routers 6 104–5
switches 6 mean time to repair (MTTR) 104
technologies 9 media access control (MAC) sub-layer
1000BaseT and 1000BaseX 30
gigabit Ethernet 11 metropolitan area network (MAN) 2
100BaseTX and 100BaseFX fast mission-critical applications 100
Ethernet 11 M/M/1/K/K queue 274–8
10BaseT Ethernet 10–11 M/M/1/K queue 271–4
10GBaseSR and 10GBaseLR M/M/1 queue 267–70
10Gbps Ethernet 11 M/M/c/K/K queue 291–6
token ring 11 M/M/c/K Queue 286–91
wireless Ethernet 11
M/M/c queue 281–6
topologies 7
multicast address 72
bus 7
multicast transmission 15–16
extended star topology 9
multipart flowspec 124–5
ring 7–8
multi-server queues 256, 281
star 7–8
M/M/c/K/K queue 291–6
transmission methods 14
M/M/c/K Queue 286–91
broadcast transmission 15
multicast transmission 15–16 M/M/c queue 281–6
unicast transmission 14–15 mutually exclusive set 160
logical link control (LLC) sub-layer 30
lognormal random variable 235–8, network analysis process 81
250–1 network architecture process 81–2
expected value and variance of network interface card (NIC) 11
237–8 network layer addresses 37
loosely coupled nodes 120 network-management protocols 90
network mask 62
MAC addresses 36–7 network-termination (NT) devices 23
Maclaurin series of exponential non-real-time applications 100
function 210 nonstationary Poisson point process
maintainability, defined 104 (NPPP) 147,150–1
management and security of network normal distribution 145–6, 153
87 normal random variable 230–5
network design goals 88 generation 250–1
number of packets in the queue and in systems and system environment

the system 260 132–4
Nyquist’s theorem 96–7 via monitoring 90
performance measures in queuing
octet 37, 63, 71–2 systems 258
OMNeT++ 130 interarrival and service time in
one-part flowspec 123–4 Poisson point process 259
Open Shortest Path First (OSPF) 45, Little’s law for queuing systems
53–4 261–2
open systems interconnection (OSI) number of packets in the queue and
reference model 28–9 in the system 260
information exchange and traffic intensity and server utilization
information formats in 31–2 260–1
interaction between OSI layers in waiting and response time 259–60
communications 30 personal area network (PAN) 2, 5
layer 1, physical layer 30 Poisson distribution 146–7, 151, 153
layer 2, data-link layer 30 Poisson point process (PPP) 145, 147,
layer 3, network layer 30 222
layer 4, transport layer 29 interarrival and service time in 259
layer 5, session layer 29 nonstationary Poisson point process
150–1
layer 6, presentation layer 29
operations of 151
layer 7, application layer 29
for queuing systems 257–8
operation, administration, and
stationary Poisson point process
maintenance applications 101
148–50
OPEX 79, 82, 88, 93
Poisson random variable 208–11, 258
optical fiber 9–10
Primary Rate Interface (PRI) 24
outcome, defined 159
private IP addresses 64
probabilities in performance analysis
packet error rate (PER) 95–7 159
packet-switched networks (PSNs) 22, axioms and properties of 161–4
25 basic concepts and terminology
packet switching 21, 46 159–61
parallel computing system 121 conditional probability 164–8
distributed computing model with independence of events 168–70
126 probability density function (pdf)
partial mesh topologies 20 188–90, 259
Pascal distribution 146 of continuous uniform random
Pascal random variable 205–8 variable 218
Pascal random variable generation 249 of Erlang random variable 226
peer-to-peer (P2P) model 118, 126 of exponential random variable 222
performance evaluation 129 of lognormal random variable 235
analytical modeling 131–2 of normal random variable 230
benchmarking 129–30 probability distribution identification
simulation 130–1 153–4
Index 321
probability mass function (pmf) 177–8, discrete uniform random variable

198, 200, 203 generation 245
of Bernoulli random variable 198–9 exponential random variable
of binomial random variable 200 generation 247
of discrete uniform random variable geometric random variable
211 generation 243–4
of geometric random variable 203 transformation techniques 248
of Pascal random variable 206 binomial random variable
of Poisson random variable 208 generation 248
probability measure 161, 175 Erlang random variable generation
protocol data unit (PDU) 31–2 250
public data networks (PDNs) 26 lognormal random variable
generation 250–1
Quality of Service (QoS) 95 Pascal random variable generation
quantile–quantile (Q–Q) plot 154 249
queuing system 141 random variables 175–6
simulation phases in 142 continuous random variables 186
statistical models in 145–6 cumulative distribution function
queuing theory 147, 255–6 187–8
computer and communication expected value 190
networks, queuing systems in moments 191–2
256–7
probability density function
Kendall’s notation for queuing
188–90
systems 262–3
variance and standard deviation
performance measures in queuing
190–1
systems 258
discrete random variables 176
interarrival and service time in
cumulative distribution function
Poisson point process 259
178–80
Little’s law for queuing systems
expected value 180–2
261–2
number of packets in the queue moments 185–6
and in the system 260 probability mass function 177–8
traffic intensity and server variance and standard deviation
utilization 260–1 182–5
waiting and response time 259–60 real-time applications 100
Poisson point process (PPP) for reliability 95, 100
queuing systems 257–8 definition 104
reliability, maintainability, and
random variable generation in network availability analyses 104–9
simulation 241 remote networks 47
inverse transform technique 241 repeaters/hubs 38–40
Bernoulli random variable requirement analysis 93, 102
generation 242–3 application-requirement analysis
continuous uniform random 100–1
variable generation 246 host-requirement analysis 101–2
network services and requirements Simple Network Management Protocol

93–4 (SNMP) 35, 89–91
reliability, maintainability, and simple repeaters 39
availability analyses 104–9 simplicity versus functionality 82
requirement-analysis model 103–4 simulation, merits and demerits of
requirement-analysis process 98–9 130–1
service characteristics 94–7 simulation model 137
user-requirement analysis 99–100 deterministic 139
requirement-analysis model 103–4 dynamic 139
requirement-analysis process 98–9 for emulating and evaluating a
response time, waiting and 259–60 system 137–8
ring topology 7–8 example 140–1
Riverbed Modeler 130 static 139
routers 6, 40 steps/phases in 140
routing in computer networks 45 stochastic 139
border gateway protocols (BGPs) of a system 139–40
56–7 single-path routing algorithms 47
dynamic routing protocols 48–9 single-server queues 256, 267
fundamentals 45 components in 134
path determination 46 M/M/1/K/K queue 274–8
routing algorithms 46–8 M/M/1/K queue 271–4
interior gateway protocols (IGPs) 49 M/M/1 queue 267–70
distance-vector routing protocols skewness 186
(DVRPs) 49–52 specialized equipment 101
enhanced interior gateway routing split horizon rule 52
protocol (EIGRP) 56 standards organizations 38
link-state routing protocols star topology 7–9
(LSRPs) 52–6 static routing algorithms 47
static routing protocols 48 static routing protocols 48
Routing Information Protocol (RIP) 50 static simulation models 139
routing loops 51–2 stationary Poisson point process (SPPP)
147–50
sample space 159 statistical models 145
security requirements 102 input modelling 152
servers 101, 120, 256 data collection 152–3
server utilization 260, 270, 274, 296 distribution fitness test 155
service-based networking 93 parameter estimation 154–5
service characteristics 94–7 probability distribution
service data unit (SDU) 31 identification 153–4
service planning 115 Poisson point process (PPP) 147
set, defined 160 nonstationary Poisson point
Shannon’s theorem 97 process 150–1
shielded twisted-pair (STP) 9–10 operations of 151
Simple Mail Transfer Protocol (SMTP) stationary Poisson point process
35 148–50
Index 323
in queuing systems 145–6 Transmission Control Protocol (TCP)

statistical multiplexing 25 34
steady-state distribution of the Transmission Control Protocol/Internet
birth–death process 261 Protocol (TCP/IP) 1
steady-state probability 268 TCP/IP suite 33–4
stochastic simulation models 139 triangular distribution 153
subnet mask 62–3 Trivial FTP (TFTP) 35
subnetting 63, 66–8 troubleshooting 83, 90
variable-length subnetting 69–70 truncated normal distribution 145
subscriber network interface (SNI) 26 two-part flowspec 124
supernetting 63, 70–1 Types of Service (ToS) 95
switched multimegabit data service
(SMDS) 26 unicast address 72
switches 6 unicast transmission 14–15
symmetric DSL (SDSL) 27–8 unshielded twisted-pair (UTP) 9–10
system, defined 93 user datagram protocol (UDP) 34
system components 94 user-requirement analysis 99–100
systems and system environment 132–4 users 93
telemetry control applications 101
variable-length packets 25
teleservice applications 101
variable-length subnetting (VLSM) 63,
Telnet 35
69–70
temporary addresses versus persistent
variance and standard deviation 182–5,
addresses 63
190–1
10BaseT Ethernet 10–11
VC packet switching 21
10GBaseSR and 10GBaseLR 10Gbps
very high-data-rate DSL 28
Ethernet 11
terminal adapter (TA) 23 virtual circuits (VCs) 20
terminals 23 PermanentVC (PVC) 22
1000BaseT and 1000BaseX gigabit Switched VC (SVC) 22
Ethernet 11 visualization applications 101
throughput, defined 6
time between failure (TBF) of a system waiting and response time 259–60
105 waiting time 259–60, 269
time to repair (TTR) of a system 105 web applications 101
token ring 9, 11 Weibull distribution 146, 153
transformation techniques 248 wide area network (WAN) 1–2, 16, 122
binomial random variable generation devices 16
248 access server 18
Erlang random variable generation channel service unit/digital service
250 unit 17
lognormal random variable modems 17
generation 250–1 multiplexers 18
Pascal random variable generation WAN switches 17
249 technologies 20
asynchronous transfer mode virtual circuits 22

(ATM) 25–6 X.25 24
circuit switching 21 topologies 18
digital subscriber line (DSL) 26–8 full mesh 19–20
frame relay 24–5 partial mesh 20
integrated services digital network wireless Ethernet 11
(ISDN) 22–4
packet switching 21 X.25 24–5
switched multimegabit data service
(SMDS) 26

Network Design, Modelling and Performance Evaluation-The Institution of Engineering and Technol

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Network Design, Modelling and Performance Evaluation-The Institution of Engineering and Technol

Uploaded by

Copyright:

Available Formats

IET TELECOMMUNICATIONS SERIES 77

Network Design, Modeling,

Volume 9 Phase Noise in Signal Sources W.P. Robins

The Institution of Engineering and Technology

The Institution of Engineering and Technology is registered as a Charity in England &

© The Institution of Engineering and Technology 2019

First published 2018

The Institution of Engineering and Technology

British Library Cataloguing in Publication Data

ISBN 978-1-78561-336-4 (hardback)

Typeset in India by MPS Limited

List of figures xiii

1 Internetworking and network fundamentals 1

2 Routing in computer networks 45

3 Internet Protocol addressing 61

4 Network analysis, architecture, and design 79

4.9 Related works 91

6 Network flow analysis 115

7 Network performance evaluation 129

8 Network simulation models 137

8.3 Steps/phases in a simulation 140

9 Statistical models in network simulation 145

10 Probabilities in performance analysis 159

11 Random variables in network modeling and simulation 175

11.5 Review questions 192

12 Discrete random variables and their applications 197

13 Continuous random variables and their applications 217

14 Random variable generation in network simulation 241

15 Queuing theory for network modeling and performance evaluation 255

15.3 Poisson point process for queuing systems 257

16 Single-server queues—network behaviors and analysis 267

17 Multi-server queues—network behaviors and analysis 281

Figure P.1 Dependence graph of chapters xxiii

Figure 1.34 X.25 devices 24

Figure 5.8 Devices connected in parallel 107

Figure 14.4 Samples of X ∼ U (1, 5) 246

Table 2.1 Dijkstra’s algorithm for finding the shortest distance

About the book

About the author

About the book

Plan of the book

Usage of the book

Figure P.1 Dependence graph of chapters

ABR area border router

EBCDIC Extended Binary Coded Decimal Interchange Code

Mbps megabits per second

STP shielded twisted pair

1.1 Fundamentals of networking

Figure 1.1 Local area network

Figure 1.2 Wide area network

Figure 1.3 Campus area network

Figure 1.4 Metropolitan area network

Figure 1.5 Home area network

Figure 1.6 Personal area network

1.2 Local area networks

1.2.1 LAN devices

Figure 1.7 Hub and its network symbol

Figure 1.8 Switch and its network symbol

Figure 1.9 Router and its network symbol

Figure 1.10 Bus topology

1.2.2 LAN topologies

Figure 1.11 Star topology

Figure 1.12 Ring topology