基于约束放松的自适应Web服务选取方法的研究  

Research of the Service Selection Method Based-on Constraint Relaxing

作　　者：刘国奇[1] 朱志良[1] 刘莹[1] 李丹程[1] 

机构地区：[1]东北大学软件学院,沈阳110819

出　　处：《小型微型计算机系统》2011年第12期2353-2358,共6页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(60872040)资助;辽宁省自然科学基金项目(20092006;20111001)资助;沈阳市科技基金项目(1091176-1-00)资助

摘　　要：已有服务选取都是把问题规划为针对某个模型寻求最优解或者次优解的过程,不考虑用户约束对Web服务选取的影响,而在SLA模型中,用户对服务的选取是有约束的,如果约束不满足,选取算法直接结束,要求用户重新给出约束定义,增加了用户与选取过程的交互.为了减少用户的操作次数,本文为用户的约束定义了约束放松模型,对于约束不能满足的情况下,自动地对已有模型进行条件放松,定义了基于分布式约束满足问题求解的服务选取过程,将约束放松模型应用于选取过程,进行自适应的选取.通过与已有的选取算法进行实验比较,本文提出的自适应组合服务选取算法,考虑了用户的约束条件,在选取失败时,基于放松模型进行自动约束条件的更新,进行服务的重新选取,减少了与用户交互次数,有效提高了有约束的组合服务选取的执行效率.The existed service selection methods consider the selection problem as getting the best or second best solution,and don′t take the influence of user′s constraint into consideration.However,in the model of SLA,user would provide his constraint on QoS.When considering user′s constraint in service selection process,if constraints are not satisfied,the selection algorithm ends and user should define constraints again.It increases the interaction times between user and selection process.In order to decrease operation times of user,the paper defines constraint relaxing model for user′s constraint.If the constraints are not satisfied,the service selection algorithm based on distributed constraint satisfaction problem would relax the constraint automatically,and selects the services adaptively.The experiment of comparing with other algorithms shows that the adaptive selection algorithm could relax the constraint automatically and re-select services adaptively;it reduces the interaction times with users,and improves the efficiency of service selection.

关 键 词：WEB服务 服务选取 自适应 约束放松模型 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]