摘要
为了提高电子商务网站资源提供者和服务提供者效用,本文提出一种新的B-PSA(双目标的池尺寸分配)算法.该算法依照排队论,参数化系统,建立以接纳概率和池尺寸为未知量的双目标优化函数,并使用启发式算法求出解;通过预测,使得各服务类的池尺寸适应变化的访问速率.与现有的算法进行模拟比较,B-PSA能提高整体的满意度.
In order to improve utility of the resource provider and service provider, A new B-PSA(Bi-criteria Pools' Sizes Allocation) scheme is presented. In this scheme, the system is modeled using parameters according to queue theory, and the Bi-criteria optimizing function that the unknown parameters are pools' sizes and admission probability is built, and the values are solved by heuristic scheme; In order to adapt to the varying arrival rate, the pool's sizes of each service class is varying by predicting parameters. Compare with the current schemes, the B-PSA increases content degree of resource provider and service providers in simulation experimentation.
出处
《小型微型计算机系统》
CSCD
北大核心
2007年第4期683-687,共5页
Journal of Chinese Computer Systems
基金
国家"八六三"重点计划项目新型网络服务器系统(2004AA111110)资助
关键词
服务级别约定
池尺寸分配
排队论
电子商务网站
service level agreements
Pools' sizes allocation
queue theory
e-commerce site