基于混合模式的流媒体缓存调度算法

发布时间：2018-06-16 14:34

本文选题：流媒体 + 补丁预取　；参考：《计算机科学》2013年02期

【摘要】：介绍了一种结合了CDN和P2P互补优势的流媒体混合内容分发网络(HyCDN)。针对HyCDN不同区域提出了相应的缓存算法,域内用户端综合考虑了流媒体前缀字节的有用性、文件的传输代价及点播热度,在此基础上提出缓存替换算法(Comprehensive Value Cache Replacement Algorithm for P2P,CVCR4P2P);对域间边缘服务器采用补丁预取与调度算法(Dynamic Scheduling Algorithm for Proxy Caching,DSA4ProxyC),通过基于用户访问情况自适应伸缩缓存的分配方案,使流媒体后缀部分在边缘服务器中缓存的数据段与其流行度成正比。理论分析及实验结果表明,混合流媒体缓存调度策略的实施能有效地降低骨干网络带宽资源消耗,对用户请求到达速率的变化具有良好的适应性。
[Abstract]:This paper introduces a hybrid streaming media content distribution network (HyCDN) which combines the complementary advantages of CDN and P2P. This paper proposes a cache algorithm for different regions of HyCDN. The client in the domain considers comprehensively the usefulness of prefixed bytes of streaming media, the transfer cost of files and the heat of on-demand seeding. On this basis, a cache replacement algorithm, Comprehensive value Cache replacement Algorithm for (P2P), CVCR4P2PU, and a patch prefetching and scheduling algorithm, dynamic scheduling Algorithm for proxy CachingDSA4ProxyCan, are proposed for inter-domain edge servers, and the allocation scheme of adaptive scaling cache based on user access is proposed. The stream media suffix is directly proportional to the popularity of the data segment cached in the edge server. Theoretical analysis and experimental results show that the implementation of the mixed streaming media cache scheduling strategy can effectively reduce the bandwidth consumption of backbone networks and has a good adaptability to the change of user request arrival rate.
【作者单位】：华侨大学计算机科学与技术学院;西安电子科技大学机电工程学院;南京理工大学计算机科学与技术学院;
【基金】：厦门市科技局科技计划创新项目(3502Z20103027) 华侨大学科研启动基金项目(09BS514) 中央高校基本科研业务费专项资金项目(JB-ZR1130) 华侨大学科研基金项目(11HZR05)资助
【分类号】：TP393.02;TN919.8

【共引文献】