最小化出口流量花费的接入级P2P缓存容量设计方法  

A P2P Cache Capacity Design Method to Minimize the Total Traffic Cost of Access ISPs

作　　者：翟海滨[1] 张鸿[1] 刘欣然[1] 王勇[1] 沈时军[1] 李正民[1] 

机构地区：[1]国家计算机网络应急技术处理协调中心,北京100029

出　　处：《电子学报》2015年第5期879-887,共9页Acta Electronica Sinica

基　　金：国家973重点基础研究发展规划(No.2011CB302605)

摘　　要：接入级P2P(Peer-to-Peer)缓存容量设计回答在接入级ISP(Internet Service Provider)出口部署多大容量缓存的问题,目前还没有最优P2P缓存容量设计方法被提出.本文提出一种权衡存储成本和带宽成本的P2P缓存容量设计方法,以最小化ISP出口流量总花费为目标,将最优缓存容量设计问题描述为整数规划问题,其目标函数形式为单调阶梯函数,通过理论推导得出最优缓存容量计算公式指导接入级ISP进行缓存容量设计.将本文所提方法与Median和"20-80 Rule"等几种ISP常用的容量设计方法进行性能比较,结果表明,本文所提方法明显优于已有方法,与目前ISP最认可的"20-80 Rule"相比,应用本文所提方法的ISP出口流量总花费最多可降低7.5%.By deploying Peer-to-Peer （P2P） caches at the edge of access ISPs, cross-ISP P2P Iraffic can be reduced. The problem is how to design an optimal cache capacity？ Up to now, no optimal P2P cache capacity design method has been proposed yet. In this paper, an optimal P2P cache capacity design method is proposed based on the storage versus bandwidth cost tradeoff. In order to minimize the total traffic cost of access ISPs, the cache capacity design problem is formulated into an integer programming problem whose objective function is piecewise continuous. The formula of optimal cache capacity can be derived through theoretical derivation.Experimental results suggest that compared with several simple design rules including No Cache,Median and ＂20-80 Rule＂, ISPs can achieve significant cost saving using our method. For example, compared with ＂20-80 Rule＂ that is most commonly used by ISPs nowadays,the cost saving can be as much as 7.5%.

关 键 词：P2P缓存技术 ISP网络 缓存部署 容量设计 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]