Self-similar network traffic characterization through linear scale-invariant system models

Abstract

It has been empirically documented that data traffic over networks of various types exhibits fractal or self-similar behavior in many instances. Accurate analysis of traffic density and estimation of buffer size must take into account this self-similar nature. There is ongoing research on generating self-similar data for use in simulation and modeling of network traffic. This paper demonstrates that the novel models proposed by Zhao and Rao (1998, 1999) for constructing purely discrete-time self-similar processes and linear scale-invariant (LSI) systems lend themselves to the synthesis of data whose self-similarity parameters match those observed in network traffic. The paper provides theoretical development and experimental results.