A Video Tra c Model based on theShifting - Level Process : the E ects of SRD andLRD on Queueing

| Recently, a number of empirical studies have demonstrated the existence of long-range dependence (LRD) or self-similarity in VBR video traac. Since previous LRD models cannot capture all short-and long-term correlation and rate-distribution while still retaining mathematical tractability, there exist many doubts on the importance of SRD, LRD, and rate-distribution on traac engineering. In this paper, we present a video traac model based on the shifting-level (SL) process with an accurate parameter matching algorithm for video traac. The SL process captures all those key statistics of an empirical video trace. Also, we devised a queueing analysis method of SL/D/1/K, where the system size at every embedded point is quantized into a xed set of values, thus name quantization reduction method. This method is diierent from previous LRD queue-ing results in that it provides queueing results over all range not just an asymptotic solution. Further, this method provides not only the approximation but also the bounds of the approximation for the system states and thus guarantees the accuracy of the analysis. Especially, we found that for most available traces their ACF can be accurately modeled by a compound correlation (SLCC): an exponential function in short range and a hyper-bolic function in long range. Comparing the queueing performances with C-DAR(1), the SLCC, and real video traces identiies the eeects of SRD and LRD in VBR video traac on queueing performance. Keywords|VBR video traac model, shifting-level process (SL), autocorrelation, long-range dependence (LRD), short-range dependence (SRD), queueing analysis.