网格服务环境下的全局最优服务选择模型研究

Global Optimal Service Selection Model Based on Grid Service

在大量相似服务共存竞争环境下,面向服务质量的Web服务选择成为研究热点.已有的服务选择方法通常假设同一时刻请求某种功能服务的请求者只有一个,并为其选择一个局部最佳服务.然而在网格服务环境下,同一时刻存在相对固定数目的多个服务提供者和多个服务请求者,要使网格服务资源利用率达到最优,需要进行整体规划,制定最优选择方案.本文利用层次分析法计算出请求者对各服务的QoS满意度与服务信任的满意度,将二者结合起来,计算出综合满意度,并基于综合满意度进行0-1整数规划,建立全局最优网格服务选择模型.最后,通过仿真实验,验证了该方法的有效性.

Web service selection based on quality of service （OoS） has been a research focus under the environment with lots of similar service existing. The current methods of service selection usually assume that there is only one service requester requesting a type of functional service at a time and choose the best one for the present requester. But actually there are many service providers and also many service requesters requesting services at a time under grid service. Therefore, it is necessary to project integrally and work out a global optimal service selection in order to optimize the resource of grid service. This paper uses AHP to calculate the satisfactions of service＇ s QoS and trust, and Combine the two satisfactions to calculate the synthesized satisfaction. Then make 0-1 integer programming based on synthesized satisfaction and work out the global optimal model for grid service selection. Finally, the effectiveness of the method has been verified by simulation.