This paper is devoted to the modeling of long-range dependent telecommunication traffic and a study of its characteristics. It consists of the following parts. Chapter 2 covers background information about the measurement of the longrange dependent and the self-similar traffic, their mathematical definitions and properties. Chapter 3 will present modeling issues of long-range dependent traffic and the approach that is used in this paper. Chapter 4 will discuss characteristics of the ARIMA(0,d,0) process, which is a fundamental self-similar process that possesses long-range dependence. Chapter 5 gives the algorithm of parameter estimation for modeling and will show our complete modeling approach using ARIMA(p,d,q) process based upon the ARIMA(0,d,0) process. Chapter 6 is about video trace data modeling using the method presented in the previous chapters. Chapter 7 will be a simulation performance study of the real video traffic and the simulated traffic. Chapter 8 concludes this paper

Analytical modeling of traffic sources in a packet-switched communications network is the basis of performance analysis and network design. Along with the development of high speed networks, more types of traffic are being applied to the network, such as real time video traffic and multimedia traffic with voice, data, image and video. To support quality service for various types of traffic, it is very important to characterize the properties of the traffic through accurate modeling. For decades, traffic modeling has been an active research area.