基于非合作博弈效用最优化的云资源提供策略  

Cloud Resource Provision Strategy Based on Non-cooperative Game Utility Optimization

作　　者：张小庆[1] 贺忠堂 李春林[1] 钱琼芬[1,3] 张恒喜[1,4] 

机构地区：[1]武汉理工大学计算机学院,武汉430063 [2]广东电子工业研究院,广东东莞523808 [3]空军雷达学院四系,武汉430019 [4]空军勤务学院,江苏徐州221000

出　　处：《计算机工程》2013年第7期59-62,72,共5页Computer Engineering

基　　金：国家自然科学基金资助项目(60970064;60773211;61171075)

摘　　要：为解决云计算资源提供过程中用户的异构性需求问题,提出一种基于非合作博弈效用最优化的云资源提供策略。利用比例共享机制,根据用户的出价提供资源,对用户的出价函数进行求解,并证明效用最优化模型存在非合作博弈纳什均衡解。实验结果表明,该策略能够反映用户需求与资源价格之间的浮动关系,规范用户的出价与资源分配,在公平性、均衡性和合理性上均有较好的效果。To solve the problem of heterogeneity of user requirements in cloud resource provision, this paper proposes a cloud resource provision strategy based on non-cooperative game utility optimization. The strategy provides resources according to user＇s bidding by proportional sharing mechanism. The bidding function is solved, and the existence of Nash equilibrium solution of optimal bidding set is proved. Experimental results show that the strategy can reflect the floating relation between user requirement and resource price, regulate the user＇s bidding and resource allocation, and get better performance on fairness, equilibrium and rationality.

关 键 词：云计算 效用最优化 资源提供策略 非合作博弈 纳什均衡 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]