摘要
P2P VoD应用中,用户行为不同步且允许随时离开系统降低了资源相互共享的概率,导致缓存可用性及服务效率降低.该问题严重影响了用户体验的质量及系统的可扩展性.定义并分析了缓存片段的相对可用性;基于对节点生存期分布规律的考察,提出一种评估缓存片段相对可用性的实用方法;设计出一种分布式的缓存优化策略:利用各节点对局部范围内缓存片段相对可用性的评估、优化来提升整个系统的缓存可用性及服务效率.仿真结果表明,该策略有效提高了播放连续度,降低了服务器负载,提高了系统可扩展性.
In P2P VoD applications,due to asynchrony in users' behavior and random peer departure,sharing ratio between peers are really low,leading to a severe impact on cache availability and its service efficiency.This problem seriously affects user experience and system scalability.Firstly,we define and analyze the relative availability of cached video clips.Secondly we propose a relative availability evaluation method according to the observation of the peers' lifespan distribution.Finally,we design a distributed cache optimization strategy:let every peer evaluate and optimize relative availability of cached video clips in local scope in order to increase cache availability and service efficiency of the whole system.Simulation reveals that this strategy increases the playing continuity,alleviates the server load and improves system scalability.
出处
《小型微型计算机系统》
CSCD
北大核心
2010年第8期1473-1477,共5页
Journal of Chinese Computer Systems
基金
国家自然科学基金项目(60773193)资助
武汉市青年科技晨光计划项目(200850731351)资助