Title :
Characterisation of video and generation of statistically equivalent Markov modulated Poisson processes
Author :
Cosmas, J.P. ; Odinma-Okafor, A.
Author_Institution :
Queen Mary Coll., London, UK
Abstract :
Determines the cell interarrival time distribution of a video codec, a number of these sources are then superposed into an aggregate cell arrival process and this stream is finally approximated to a statistically equivalent Markov Modulated Poisson Process (MMPP). The MMPP representations of video sources can then be superposed to generate composite sources. The motivation for generating an equivalent MMPP is to facilitate theoretical analysis of queueing systems performance and for generating composite sources for simulation studies. Theoretical analysis of queueing systems performance is important since it yields important measures of cell loss and transfer time at the extreme ends of their distributions. This work uses the techniques of Heffes and Lucantoni (1986)
Keywords :
Markov processes; codecs; modulation; queueing theory; video signals; Markov modulated Poisson processes; cell arrival process; cell interarrival time distribution; composite sources generation; queueing systems performance; video codec; video sources;
Conference_Titel :
Packet Video, IEE Colloquium on
Conference_Location :
London