一种Multi-Agent的云联盟服务组合模型  

Cloud Federation Service Composition Model Based on Multi-Agent

作　　者：尚广安 蒋哲远[1] 

机构地区：[1]合肥工业大学计算机与信息学院,安徽合肥230009

出　　处：《微电子学与计算机》2014年第9期1-6,共6页Microelectronics & Computer

基　　金：国家自然科学基金重点资助项目(60633060)

摘　　要：提出了一种利用Multi-Agent的交互协作以实现云联盟平台中的云服务自动优化组合模型.继而基于该模型设计了一种启发式云服务组合算法,该算法首先根据用户的个性化请求对服务库中的云服务进行预处理,筛选出合适的云服务并对其进行代价遍历.最后,以服务相似度和组合QoS对选取的云服务进行回溯优化组合,进而获取最优的组合方案.仿真结果表明该算法在组合效率和成功率方面较同类算法有较大提升.A cloud federation service composition model based on Multi-Agent is proposed to compose optimal services in multi-Cloud platform.Then on the basis,a new heuristic algorithm is designed.According to the input parameters of request,the cloud service in the service library is preprocessed.And then the algorithm traverses the appropriate cloud service with heuristic function.Finally,the algorithm makes full use of service similarity and QoS to choose concrete service and program optimized in the process of regression.Compared with other similar algorithms,the simulation experiments show our algorithm can provide higher efficiency and shorter solution path for the requests.

关 键 词：云联盟 MULTI-AGENT 服务组合 启发式算法 

分 类 号：TP393.09[自动化与计算机技术—计算机应用技术]