An analytically tractable model for video conference traffic

In this letter, we propose an analytically tractable approach to model-comp ressed video traffic called C-DAR(1), The C-DAR(1) model combines an approa ch utilizing a discrete-time Markov chain with a continuous-time Markov cha in. We show that this approach accurately models the distribution and expon ential autocorrelation characteristics of video conferencing traffic. Also, we show that by comparing our analytical results against a simulation usin g actual video-conferencing data, our model provides realistic results. In addition to presenting this new approach, we address the effects of long-ra nge dependencies (LRD) in the video traffic. Based on our analytical and si mulation results, we are able to conclude that the LRD have minimal impact on videoconference traffic modeling.