基于分簇结构的P2P流媒体混合分发算法  

A HYBRID SCHEDULING ALGORITHM FOR P2P STREAM BASED ON CLUSTERING

作　　者：陈瑞昭[1] 刘永广[1] 

机构地区：[1]广东轻工职业技术学院管理工程系,广东广州510300

出　　处：《华南师范大学学报（自然科学版）》2010年第1期37-41,共5页Journal of South China Normal University(Natural Science Edition)

基　　金：广东省计算机网络重点实验室开放基金资助项目(cn200407);广东轻工职业技术学院自然科学基金资助项目(KX010405)

摘　　要：提出了一种基于分簇结构的混合分发算法,算法采用分簇的方法将流媒体中的节点资源进行簇划分,形成由簇头、簇内节点构成的分簇网络结构,簇头与簇内节点通过拉拽算法来获得数据,而簇头间采用推送分发算法.仿真结果表明,该算法能提高数据块复制速度,减少数据传播时延,有效降低系统的控制开销,提高了播放连续度.A hybrid scheduling algorithm based on clustering is presented. In the algorithm, node resources are divided into cluster head and cluster nodes by clustering method and form a clustering network structure. For the network, data is obtained by push algorithm between cluster heads and heads, while pull algorithm between cluster heads and inter-cluster nodes. Simulations show that the algorithm can increase the speed of data block duplication, reduce the data propagation delay and the control overhead of streaming system efficiently, and improve the degree of continuous playback.

关 键 词：对等网络 流媒体 P2P覆盖网 簇 推送算法 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]