一种分阶段的P2P流媒体数据调度算法

针对当前P2P流媒体中各节点数据调度时独立获取数据而对提高节点间数据协作性考虑不足，导致服务器负载较重的问题，提出一种分阶段的数据调度算法。当多个节点将请求的数据中含有相同数据块时，各节点将分两个阶段合作的获取这些数据。第1阶段按照带宽相对大小各节点以合作的方式先各自获取数据的一个子集，第2阶段节点间再充分共享上一阶段已经获得的子集数据，协作地完成该轮调度的数据传输。仿真实验表明该算法可以提高节点间数据协作性，充分利用节点带宽资源传输媒体数据，提高系统可扩展性。

Step based data scheduling algorithm for P2P streaming system

Existing data scheduling algorithm makes insufficient consideration to improve data collaboration among peers, which leads to heavy server load in current P2P streaming system.In order to solve this problem,a step based scheduling algorithm is proposed.When there are more than one peer will request same data chunks in one scheduling period,each peer will obtain these chunks in 2 steps.In step 1,each peer will obtain part of chunks cooperatively.In step 2,each peer will share chunks which have been obtained in step 1 with other peers.Simulation shows that the proposed algorithm can enhance data collaboration among peers and make full use of peers’bandwidth resource to transmit media chunks,which im proves system scalability.