Network Design, Modelling and Performance Evaluation
Designed for ICT professionals involved in the planning, design, development, testing and operation of network services, this book is ideal for self-teaching. It will help readers evaluate a network situation and identify the most important aspects to be monitored and analysed. The author provides a detailed step by step methodological approach to network design from the analysis of the initial network requirements to architecture design, modeling, simulation and evaluation, with a special focus on statistical and queuing models. The chapters are structured as a series of independent modules that can be combined for designing university courses. Practice exercises are given for selected chapters, and case studies will take the reader through the whole network design process.
Other keywords: network performance evaluation; internetworking; network design; network flow analysis; network fundamentals; routing; random variable generation; computer networks; performance analysis; single-server-queues; continuous random variables; Internet Protocol addressing; multiserver queues; network-requirement analysis; network architecture; network analysis; statistical models; discrete random variables; queuing theory; network simulation models
- Book DOI: 10.1049/PBTE077E
- Chapter DOI: 10.1049/PBTE077E
- ISBN: 9781785613364
- e-ISBN: 9781785613371
- Page count: 354
- Format: PDF
-
Front Matter
- + Show details - Hide details
-
p.
(1)
-
1 Internetworking and network fundamentals
- + Show details - Hide details
-
p.
1
–43
(43)
In this chapter, the author first outline the fundamentals of networking in Section 1.1 with different network 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 the author 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 protocol 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 Section 1.6 where I will show the functionalities of intermediary devices for computer networking, including repeaters/hubs, bridges/switches, and routers/gateways.
-
2 Routing in computer networks
- + Show details - Hide details
-
p.
45
–59
(15)
In this chapter, for a first view of routing in computer networks, the author will first discuss 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 protocols (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).
-
3 Internet Protocol addressing
- + Show details - Hide details
-
p.
61
–77
(17)
In this chapter, the author 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.
-
4 Network analysis, architecture, and design
- + Show details - Hide details
-
p.
79
–92
(14)
In this chapter, the author 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.
-
5 Network-requirement analysis
- + Show details - Hide details
-
p.
93
–113
(21)
In this chapter, the author 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.
-
6 Network flow analysis
- + Show details - Hide details
-
p.
115
–128
(14)
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.
-
7 Network performance evaluation
- + Show details - Hide details
-
p.
129
–135
(7)
In order to validate the effectiveness 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, the author 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.
-
8 Network simulation models
- + Show details - Hide details
-
p.
137
–143
(7)
This chapter will discuss in detail simulation models for emulating and evaluating a system. In this chapter, the author 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.
-
9 Statistical models in network simulation
- + Show details - Hide details
-
p.
145
–157
(13)
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 queuing 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 inhomogeneous 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.
-
10 Probabilities in performance analysis
- + Show details - Hide details
-
p.
159
–173
(15)
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 probabilities 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 chapter will help them review these and present the terminology and concepts in the communication networks. In this chapter, the author 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.
-
11 Random variables in network modeling and simulation
- + Show details - Hide details
-
p.
175
–196
(22)
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, the author 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.
-
12 Discrete random variables and their applications
- + Show details - Hide details
-
p.
197
–216
(20)
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 random variables in Section 12.1. Then, various kinds of discrete random variables, including Bernoulli, binomial, geometric, Pascal, Poisson, and discrete uniform random 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.
-
13 Continuous random variables and their applications
- + Show details - Hide details
-
p.
217
–240
(24)
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.
-
14 Random variable generation in network simulation
- + Show details - Hide details
-
p.
241
–254
(14)
This chapter is devoted to illustrating some widely used techniques for generating random variables. In this chapter, the author 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.
-
15 Queuing theory for network modeling and performance evaluation
- + Show details - Hide details
-
p.
255
–265
(11)
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, the author 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.
-
16 6 Single-server queues—network behaviors and analysis
- + Show details - Hide details
-
p.
267
–280
(14)
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, the author 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.
-
17 Multi-server queues—network behaviors and analysis
- + Show details - Hide details
-
p.
281
–298
(18)
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, the author 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.
-
Back Matter
- + Show details - Hide details
-
p.
(1)