基于推拉结合机制的P2P流媒体分发算法

随着流媒体直播应用的日益扩大,传统C/S（Client/Server）模式网络构架的服务器负载过重,导致QoS（Quality of Service）得不到保证,而P2P（Peer-to-Peer）网络具有负载均衡、可扩展性、容错性强等优点。目前P2P技术开始逐步走向成熟。大多数P2P流媒体数据分发采用Pull模式或Push模式,本文提出基于推拉结合机制的P2P流媒体分发算法,通过仿真实验,从系统开销、端到端延迟等方面对两者进行了性能对比。实验结果表明,该推拉结合机制有效的降低了数据块传输时延,并且在很大程度上避免了数据块的重复推送,降低了数据冗余。

P2P Streaming Media Distributing Algorithm Based on Push-Pull Mechanism

With the live streaming media application growing, the server of traditional C/S（Client/Server） mode network architecture overloads, which leads to naked QoS （Quality of Service）.P2P （Peer-to-Peer） network has advantages of load balancing, scalability, strong fault tolerance and so on.P2P technology is coming to mature. Most P2P streaming media data distribute through pull mode or push mode. This paper put forward P2P streaming media distributing algorithm based on push-pull mechanism. Through simulation experiment, both is compared on system cost, end-to-end delay performance and so on. The experiment results show that push-pull mechanism effectively reduces transmission delay of data block, and largely avoids data block duplicate push, and reduces data redundancy.