北京航空航天大学学报

• •    

实时流媒体P2P可收敛反馈网络结盟算法研究

沈孟如,张晋豫   

  1. 北京交通大学
  • 收稿日期:2015-05-12 修回日期:2015-12-13 发布日期:2016-01-20
  • 通讯作者: 沈孟如
  • 基金资助:
    国家自然科学基金;中央高校基本科研业务费专项资金

Research on the Alliance Algorithm of Converging Feedback Network on P2P Real-Time Media Streaming

  • Received:2015-05-12 Revised:2015-12-13 Published:2016-01-20

摘要: 本文提出了一个基于距离汇聚的结盟算法,在动态业务量环境中通过实现可收敛反馈网络,有效消除随机结盟、内容类似驱动结盟、带宽类似驱动结盟中存在的业务量不合理问题。实现了一个基于静态IP距离算法和动态包探测测距的混合节点间距离评估机制,较好解决了探测包测距法受业务量波动影响较大、客户端感知测距法实时性较差,以及IP包测距法误差较大的缺点。引入一个具有位置意识的基于Polling的均匀流协同机制,在保持了推-拉周期请求机制开销小优点的同时,提高对抗Serving Peer传输劣化和失败的弹性。仿真结果表明机制可以减少业务量不合理和用户不贡献恶意行为的概率,当节点较多时,机制能够提供比其它结盟机制更好的时延、丢包率和到达率性能。

关键词: 流媒体, P2P, QoS, 反馈网络

Abstract: The performance likeness alliance is a widely used method for optimizing user fairness, QoS, and network traffic in P2P media streaming system, as well as optimizing the network traffic. This paper supposes a distance-convergence alliance algorithm, by implementing the converging feedback network in dynamic traffic condition, it can avoid the unreasonable traffic which can't be eliminated by the random-alliance, the content-likeness driven alliance and the bandwidth-likeness driven alliance. This paper implements a node mix distance algorithm which combines IP static range algorithm and dynamic packet probe ranging method, it can overcome the drawbacks that packet probing range is easily affected by the traffic fluctuation, and that the client perception range can't ensure the real-time, as well as that the IP-based range algorithm lack precision. This paper introduces a polling-based collaboration mechanism of uniform flow, it not only keeps the lower-offset merit of the pushing-pulling periodic requesting mechanism, but also it provides enough resilience against to transport ability decline and transport failure. The simulation results show that our mechanism can reduce the probabilities of unreasonable traffic and the vice behaviors of non-contribution, and can provide better performance of delay, packet loss rate and packet arriving rate than other alliance mechanisms in larger scale.

Key words: media streaming, peer to peer(P2P) , quality of service, feedback network


版权所有 © 《北京航空航天大学学报》编辑部
通讯地址:北京市海淀区学院路37号 北京航空航天大学学报编辑部 邮编:100191 E-mail:jbuaa@buaa.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发