【作者】 刘丽； 方金云； 梁对；

【Author】 Liu Li;FANG Jin-yun;Liang Dui;Institute of Computing Technology,Chinese Academy of Sciences;Institute of China Electric Equipment and System Engineering Co.Ltd;China Electric Equipment and System Engineering Co.Ltd;

【机构】 中国科学院计算技术研究所计算机应用研究中心； 中国电子设备系统工程公司研究所； 中国电子设备系统工程公司；

【摘要】 云计算技术的发展使得云环境中服务资源不断增多,用户面临如何快速选择所需要服务的困难。现有的服务选择方法多是从单个用户角度出发进行的服务选择,当多个用户进行服务选择时,由于每个用户"自私"的选择最优服务易造成最优服务过载,性能下降。服务提供商作为资源的拥有者,能够对资源进行统一规划将用户引向与其QoS约束相匹配的服务上,以使得全网资源得到均衡。基于这一思想,本文从服务提供商的角度出发,使用扩展二分图对服务选择问题进行建模,将群体最优服务选择问题转换为用户与服务之间的最优匹配问题,并提出了OGSP算法来求解。OGSP算法在提高群体满意度的同时提高单个用户的满意度,在服务资源冲突时能够优先满足高等级用户的需求,同时还帮助服务提供商分散单个服务的负载,改善了资源利用率。更多还原

【Abstract】 The rapid development of cloud computing technology increases the number of cloud services,and it is difficult for users to select the required service quickly.At present,most researches select the service from the perspective of a single user.However,if there are multiple users selecting the services at the same time,the best service will be selected by too many users and the quality of the service will decrease.The cloud service providers,as the resource owners,can direct the users to the services matching their QoS constraints based on the unified planning of resources.This paper proposes an optimal group web service selection algorithm named OGSP to solve multi-users service selection problem from the prospective of cloud service providers.OGSP algorithm models the problem using the extended bipartite graph and solves it by finding the optimal matching between services and users.OGSP algorithm improves not only the group satisfaction but also the individual’s satisfaction,and the users with higher rank can get services with better QoS.To service providers,OGSP helps them to distribute the request of users,so it improves resource utilization.更多还原