Modeling and Performance Analysis of LFSN

Abstract

There is much experimental evidence that network traffic processes exhibit ubiquitous properties of self- similarity and long-range dependence (LRD). Modeling and performance evaluation of self-similar traffic is a research hotspot of computer network . The most commonly used model of self-similar traffic is fractional Brownian motion (FBM) process, which can only capture the self-similar and long-range dependent characteristics of the traffic. Recent experimental studies have shown that several traffic classes of real traffic exhibit higher variability than that captured by FBM model. In this paper, a fractional stable motion self-similar model, linear fractional stable noise (LFSN), which can capture both long-range dependent and bursty (heavy-tail) characteristics of the traffic, is studied. Based on the buffer size overflow given by the predecessors, the formulae of average queue length, queue length variance, average delay, jitter and effective bandwidth are derived. The variation of packet loss rate, average delay and effective bandwidth with the parameters of Hurst parameter, characteristic exponent and buffer size is investigated through simulation.