Modelling video telephony by Fokker-Planck equation and their analysis

Abstract

Video services Will form a substantial portion of the traffic earned by emerging broadband integrated networks. The Video streams are compressed and sent to the network to conserve bandwidth and hence require a good model to represent them In an asynchronous transfer mode (ATM) network. In this paper, we model the Video sources by means of a diffusion processes With mean fl, variance d and co-variance C( t). We consider video teleconferencing where the changes m the successive frames are not abrupt but are more or less uniform The matching parameters for the mean, variance and covariance are taken from discrete-time Markov chain (DTMC) found in Blondia and Casals (Performance Evaluation, 1992, 16, 5-20). The video telephony session is modelled according to a truncated diffusion process and we calculate the steady-state behavior for such a process We consider two truncations, namely, reflecting barrier and absorbing barrier and state their intuitions III Video telephony.