Research On Live Scheduling Algorithm In P2P Media Streaming Based On DONet

With the growing popularity of broadband and the rapid development of communication network technology, P2P media streaming that has high system scalability and good cost performance is becoming an active force to drive the applications of future broadband. How the various sender's data transmission scheduling to become effective and reasonable was a key of improving the quality of media streaming services. So a high-performance data scheduling algorithm has became a core problem in the research field of P2P media streaming.DONet is a Data-driven Overlay Networks for live media streaming and utilizes the available information of data to guide the flow of data rather than using a particular network structure to restrain it. DONet has simple, efficient, robust and other characteristics, these features improve the stability of the system and enhance the system's throughput. Thus it can effectively compensate for past deficiencies in a tree structure. However, in existing scheduling algorithm of DONet, there are many problems such as the great start delay time, data blocks easily lost under the conditions of the limited bandwidth and bandwidth utilization is not high.Aiming at media streaming data's characteristic of a time limit, on the basis of depth analysis of existing data scheduling algorithms, this paper proposed a live optimize scheduling algorithm of P2P media streaming based on DONet(LOSA).Firstly, algorithm considered the issues to associate the location of the data block in the cache of Peer node with it's broadcast time limit, and to minimize the waiting time for urgent requests, designed a strategy of urgent priority data request to ensure the prior scheduling of urgent data and reduce the losses of data blocks. Secondly, under the premise of consideration to node capabilities , algorithm combined traditional RF strategy with the proposed urgent priority data request strategies, this not only ensures the priority of urgent data requests, and also ensures the rapid distribution in network of scarce data blocks, so consequently improves the broadcast quality of media stream.Finally, the new algorithm goes through the simulation on network simulation platform NS-2, and is compared with existing data scheduling algorithms. The simulation results show that the proposed algorithm has better performance than existing data scheduling algorithms.