Subexponential Parsimonious Modeling of the Correlation Structure of VBR Video Traffic

With the increasing popularity of multimedia applications, video data represents a large portion of the traffic in modern networks. Consequently, adequate models of video traffic, characterized by a high burstiness and a strong positive correlation, are very important for the performance evaluation of network architectures and protocols. This paper presents a new model for traffic with persistent correlations based on the M/G/infin process. We use a new discrete distribution for the service time of the M/G/infin queueing system, flexible enough to give rise to a wide range of correlation structures of the resulting M/G/infin process in a parsimonious way. The corresponding simulation model is easy to initialize in steady state. Moreover, as the distribution has subexponential decay, we can apply a highly efficient and flexible generator of synthetic traces of the M/G/infin process. As an example, we analyze the capabilities of our model for capturing the empirical distribution and correlation structure of VBR MPEG video traffic.