Loss behavior of an Internet router with self-similar input traffic via Fractal point process

Abstract

It has been reported that modeling a self-similar network traffic is of key importance for the traffic engineering. Self-similarity or long range dependence causes degradation of Internet router performance. Therefore, it is decisive for an appropriate buffer design of a router. In this paper, we investigate loss behaviour of network router with pseudo self-similar traffic input. We use Fractal point process (FPP) as input process as it generates self-similar traffic. For queueing analysis, input process is Markov modulated Poission process (MMPP), which is fitted for FPP by equating the second-order statistics of counting function. The reason is, FPP is not suitable for queueing based performance evaluation. FPP involves another parameter fractal onset time (FOT) besides Hurst parameter. Effect of FOT on loss behavior is examined.