Stochastic Features of VBR Video Traffic and Queueing Working Conditions: A Simulation Study Using Chaotic Map Generator

Abstract

The paper presents a simulation study on the impact ofLRD and SRD on a simple queueing system and shows that, while for short buffers an analytically tractable SRD model may suffice to capture the long-term loss probability, for systems with large buffers the effect of LRD can be significant. The main contribution of this paper consists in highlighting this evidence through simulations driven by synthetic traces (generated by traditional AR models and LRD Chaotic Map Generator), measured data and shuffled versions of them.

Analysis of CMPP Approach in Modeling Broadband Traffic

“…The peak rate error is very high (nearly 43%) due to the very unfavorable condition on the power coefficient distribution. Indeed, the power spectral decomposition of the considered trace presents only a single couple of dominant effective real eigenvalues and, at the same time, its marginal distribution exhibits an upper tail behavior heavier than gaussian [4]. The modeling results are shown in Fig.…”

“…In the second set of simulations, two real traffic traces have been considered: the two data sets are related to a videoconference service and a LAN traffic trace. The description of the characteristics of the videoconference traffic data are described in [4], whereas the second trace is the well known "October89" trace collected at Bellcore Labs. In both analyses, the data (i.e.…”