Research On Data Scheduling Algorithm And Peer Selection Strategy Of P2P Streaming System

With the development of the network,it has been becoming a popular manner for people using streaming system to watch programs.The streaming system is gradually seizing the traditional market of cable TV.For full use of the large number of peer resources in the network,P2P-based streaming media system ensured the quality of streaming service and improved the playing performance.So,the P2P model gradually replaced the C/S model in the streaming media system nowadays,In P2P streaming systems,the live system restricted users’interactivity.So,more and more users required using Vedio-on-Demand streaming media system.Because of the asynchronous and high interactivity demand in the VoD streaming media system,it demands higher technology to implement VoD system relative to that of live system.The optimization of data scheduling strategy can improve the continuity of playing and reduce the startup latency.The improvement of peer selection strategy can enhance the quality of streaming service and optimize the performance of the system.It highlighted the research in industry and academia to improve the data scheduling and peer selection strategy.The main contributions and contents of this article:First,this article proposed a multi-level feedback queue style data scheduling strategy which based on peer dynamic bandwidth adaptive.This strategy considered the high dynamic,heterogeneous,real-time data requirement etc.characters of the peers,and select appropriate piece of data and scheduling strategy according to the dynamic change of peer service capacity and flexible time difference of data pieces.It reduced insignificance bandwidth consumption and improved the instantaneity and continuity of the system.Secondly,this article divided the peer buffer into different parts in different stages of playing.In the initial stages of playing,it loads upcoming data in the emergency area to reduce the loading time.In the stable stages of playing,it loads several videos that have been played in the played area.In this way,it will increase the service range of the peers,reduce the load on the server and thereby reduce operating costs.Finally,this article proposed an unknown overlay network topology peer selection strategy.According to the highly dynamic character of P2P streaming media system peers and non-normal exit of cooperation peer,it used routing similar thoughts to detect the area which the peer belongs to.It also dynamically assessed the service capacity of cooperation peer and selected the service peer set in the basis of historical information.In this way,it reduced the packet loss rate and enhanced the quality of streaming service.