Improvement Of The P2P Live Streaming System Based On Simulation

Recently, streaming applications based on Peer-to-Peer networks has became an imporatant application on the Internet, at the same time the large scale deployment of many P2P Live Streaming systems and Video On Demand systems has aroused intensive interest among researchers. Take PPLive as an example, at present the number of its users all over the world has exceeded one hundred million, and the number of real time online users has exceeded two million. Also, PPLive consumes 10 percent of the total bandwidth on the backbone network in China. Compared with the Live Video systems based on traditional Clients-Server mode, P2P Live Streaming systems can provide higher viewing quality for more users.In this paper, we introduce the basic principle of P2P Live Streaming system and some key technologies in the system. Through simulation, we simulate the working process of the P2P Live Streaming system, and we research the load capability of the system. We also analyze the process of the spreading of media pieces over the entire network from the perspective of single media piece. In order to shorten the spreading time of media pieces to different peers and improve the users'playback continuity, we put forward an self-adaptive policy-"+1/-1 Policy", which improves the utilization of bandwidth in the streaming system, especially enables high capacity peers to fully utilize their upload bandwidth and accelerates the spread of media pieces. To alleviate the server's load in the initial process of pieces'spread, we put forward a "Limit the Number of Pieces Policy", which accelerates the injection of new pieces into the network. We refer to the ideas in related papers, and implemented the "Server Chooses Neighbors Policy" and the "Delaying Request by Low Capacity Peers Policy" with our simulator to validate them and compare them with our policies.