Internet Traffic Modeling using the Index of Variability

In this paper, we propose a novel and mathematically rigorous measure of variability, called the index of variability (Hv ), that fully and accurately captures the degree of variability of a typical network traffic process at each time scale and is analytically tractable for many popular traffic models. Using this proposed measure, we then analyzed two traffic models: the Two-State Markov Modulated Poisson Process (MMPP) and the renewal process with hyperexponential interarrival time distributions of order two (RPH2). Two-state MMPP models are popular in modeling the superposition of packet voice streams. The results show that the traffic variability can exhibit a nonmonotonic behavior. In addition, the results suggest that renewal processes with interarrival times hyperexponentially distributed are suitable for modeling network traffic processes with high variability over a broad range of time scales.