OPTIMIZED VIDEO STREAMING USING MIXED BAYESIAN OPTIMIZATION TECHNIQUE IN P2P VIDEO ON DEMAND

Peer-to-peer (P2P) streaming tries to achieve scalability and at the same time it meets real-time playback requirements. The performance of the peer-to-peer streaming systems is the streaming quality they can provide to the peer for the better performance. It depends on the structure of the overlay as well as the way of peers exchange data with each other peers. In this paper, we generate a model based on the Mixed Bayesian Optimization Algorithm that can be used to compare different downloading strategies to random peer selection. We first study two simple strategies: Rarest First (RF) and Greedy. The former technology is a well-known strategy for P2P file sharing that gives good scalability by trying to transmit the chunks of a file to as many peers as quickly as possible. The next strategy is an intuitively reasonable strategy to get urgent chunks first to maximize playback continuity from a peer's local perspective. Next we consider a number of numerical examples on both discrete and continuous problems to illustrate our results and their application to protocol design. Finally we validate our model with simulation and show that our proposed system outperforms the conventional approaches.