一种媒体直播的P2P自适应网络快速传输模型  

A Model of Adaptive Fast-forward Control on P2P System Towards Live Streaming

作　　者：周永福[1] 曾志[2] 黄日胜[1] 

机构地区：[1]河源职业技术学院电子与信息工程学院,河源517000 [2]惠州学院计算机科学系,惠州516007

出　　处：《计算机与数字工程》2015年第6期1053-1057,1068,共6页Computer & Digital Engineering

基　　金：河源市社会发展科技项目(编号:2013-113);博士启动基金项目(编号:2009C33011)资助

摘　　要：为解决媒体流在网络直播的低延迟与可扩展性,结合P2P的特点,研究了一种基于P2P的自适应网络快速传输模型。该模型首先对网络节点状态进行分类,通过一种位索引机制实现节点状态的识别,接着在MST覆盖网络基础上,探讨了一种面向请求节点构建MST的P2P媒体直播系统,并提出了一种自适应网络快速传输控制算法。数学模型分析与模拟结果表明提出的模型能较大程度地缩短端到端传播的延迟时间,且减少跳点数与服务负载,从而进一步提升了媒体播放的流畅性,为采用P2P技术实现媒体的快速直播提供了很好的参考与借鉴。To address the problem of low delay and scalability on media living streaming in internet,combing the characteristic of P2 P,this paper investigates a model of adaptive fast-forward on P2 Psystem.Firstly,the model classifies peers according to its status,then achieves the peers status by identifying mechanism on bit index.Based on minimum spanning tree（MST）overlay,a media living system using request peer to construct MST is discussed,also an algorithm of adaptive fast-forward control is proposed.The analysis of mathematical model and simulation results show that our model can widely shorten the delay between ends and reduced hops count and server load so as to further promote smooth for media playing.Similarly,our model can provide a valuable reference for implementing media fast living on P2 Psystem.

关 键 词：流媒体 P2P 最小生成树 传输控制 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]