This book develops analytical methods from first principles to evaluate performance measures of computer systems and networks. It attempts to strike a balance between mathematical skills, generality, rigor, focus, and model formulation for application systems in computer networks. Calculus, matrix algebra, and elementary probability are prerequisites.
The book is organized into ten chapters and an appendix that reviews probability theory. Chapter 1 describes activities in computer systems and networks resulting in various types of queues, to motivate the students. Chapter 2 is devoted to traffic models, Pareto distribution, simulation, parameter estimation, and mean square convergence. Chapter 3 considers an M/M/1/∞ queue in detail. Chapter 4 deals with state-dependent Markovian queues and their applications in carrier sense multiple access with collision detection (CSMA/CD), Aloha, and local-area networks. Chapter 5 discusses an M/G/1 queue with an application to voice-over-Internet protocol (VoIP) using the Poisson arrivals see time averages (PASTA) property. Chapter 6 covers discrete time queues, chapter 7 analyzes Markovian queueing networks, and chapter 8 considers the G/M/1 queue. The effective load as a function of normalized load and the Hurst parameter of the Pareto interarrival times are illustrative. Chapter 9 introduces bursty traffic models and their effects on queues, and presents a tractable approximation to self-similar traffic. Chapter 10 deals with fluid-flow models.
Many concepts of probability theory and stochastic processes are explained with the help of queues as applications. This approach gives students motivation to study the needed principles and results. The book uses alternative and simpler techniques without using concepts from higher mathematics. This avoids undue generality and keeps the focus on the necessary results.
The book contains a wide variety of queueing models, including bursty, Markov modulated Poisson process arrivals, approximate self-similar traffic, and fluid flow. It provides a brief yet rigorous, self-contained review of elementary probability theory in the appendix. The book can serve as an ideal text for undergraduate courses in computer science and communication. Being self-contained, it can also serve as a reference for practicing engineers.