摘要
P2P流媒体直播系统主要涉及成员管理和数据调度等两方面研究。本文主要研究数据调度问题,并提出了基于最大流的分布式数据调度算法。对于每个节点,首先将节点缓冲区分为紧急区域和非紧急区域两部分,然后分别将两个区域的数据调度问题转为最大流问题,最后根据最大流问题的求解结果给出该节点的数据调度策略。该数据调度算法既考虑了数据块的时间限制的特性,又能充分利用节点的带宽,增加系统的负载平衡。利用P2PStrmSim仿真器仿真,结果表明本文所提出的分布式数据调度算法的性能比传统调度策略有显著提高。
：The research mainly involves two aspects of member management and data scheduling in P2P live media streaming system. In this paper, we study the problem of data scheduling, and propose a distributed data scheduling algorithm based on maximum flow. For every node, the buffer area is divided into emergency area and nonemergency area. Then, the data scheduling problem is transformed into a maximum flow problem in the two areas respectively. Finally, according to the result of maximum flow problem, we give a data scheduling strategy for the node. This algorithm not only considers the time limit of data blocks, but also takes advantage of the band- width of the node. By P2pStrmSim, the simulation results show that the performance of the proposed distributed algorithm is superior to some traditional data block scheduling strategies.
出处
《网络新媒体技术》
2014年第4期24-29,共6页
Network New Media Technology
基金
国家高技术研究发展计划(863)资助课题(NO:2008AA01A317)
国家自然科学基金(NO:71171189
NO:10831006
NO:10671024
NO:11331012)
