基于蚁群算法的Web集群文档优化分布方案  

Document optimization distribution strategy based on ant colony algorithm for web server cluster

作　　者：熊智[1] 熊步云[1] 周寅聃[1] 

机构地区：[1]汕头大学计算机科学系,广东汕头515063

出　　处：《计算机工程与设计》2009年第14期3389-3392,3395,共5页Computer Engineering and Design

基　　金：广东省自然科学基金项目(07301411)

摘　　要：在Web集群中优化分布海量级的Web文档是一个急需解决的问题。提出了一种以减少系统平均响应时间为目的的Web集群文档优化分布方案。该方案合适地拷贝网页簇,并通过对服务器进行建模将网页簇的分布问题转化为0-1整数规划问题。针对该问题的特点,设计实现了一种基于蚁群算法的求解方案,算法中蚂蚁对路径的选择分两步进行,并设置合适的启发值以加快收敛速度。实验结果表明了应用蚁群算法求解Web集群文档优化分布问题的可行性与有效性。How to distribute the mass documents in web server cluster is an urgent problem to be solved. A strategy to distribute web documents in web server cluster is proposed, whose aim is to reduce system＇ s average response time. It copies webpage clusters properly, and translate the document optimization distribution problem into a 0-1 integer programming problem by modeling web servers. Aimed at the traits of such problem, a solving method based on ant colony algorithm is designed and implemented, in which ants choose path in two steps, and heuristic values are properly set to fasten convergence speed. Example results verify the feasibility and effectiveness of solving document optimization distribution problem by ant colony algorithm.

关 键 词：WEB集群 文档优化分布 0-1整数规划 蚁群算法 服务器建模 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]