Markov Chain-Based Reliability Modeling of Multi-Stream P2P Video on Demand Systems

Multi-stream P2P video has been one of the contenders for the technology to be used for next generation video on demand systems which are expected to support high bit rates and long video lengths. One of the problems with multi-stream video is peer churn that is known to have an adverse effect on all proposed P2P systems including the ones involving multiple streams also known as P2P swarming. To mitigate the churn effect, various sub-stream encoding techniques with error control have been proposed for multi-stream video. In this paper, we introduce a simple continuous-time Markov chain-based stochastic model to describe peer churn and the role of error control. Using this model, we quantify the role of error control on system performance so as to be used for engineering next generation multi-stream P2P video on demand systems.