一种云间合作博弈资源提供方法  

RESOURCE SUPPLY METHOD OF COOPERATIVE GAME AMONG CLOUDS

作　　者：董玮 Dong wei(The Open University of Jilin,Chuangchun 130022,Jilin,China)

机构地区：[1]吉林开放大学,吉林长春130022

出　　处：《计算机应用与软件》2021年第3期39-45,共7页Computer Applications and Software

基　　金：吉林省科技发展计划资助项目(20190902010TC);吉林广播电视大学校级课题(2018YB07)。

摘　　要：为了满足云资源消费者的需求,并有效扩展云资源的服务能力,设计基于云间合作博弈的资源联盟形成算法。以联盟总体利益最大化为目标,将多个云资源提供者间的合作行为建模为联盟博弈模型,从而得到最优联盟结构满足用户方的虚拟机实例请求;设计一种联盟的合并与分裂规则,使得最终联盟结构是稳定不变的;设计基于联盟成员贡献能力的标准化估计Banzhaf值法实现联盟总体利益的个体分割。实验结果表明,在不同虚拟机请求类型测试条件下,云联盟形成算法不仅可以确保更高的联盟总体利益,而且使利益分割更加公平,其算法执行效率也更高。In order to satisfy the requirements of cloud resource consumers,and expend the server capacity of cloud resources effectively,a resource providers coalition formation algorithm based on cooperative game among clouds is presented.With the goal of maximizing the overall profit of coalitions,my algorithm modelled the cooperative behaviors among multiple cloud resource providers as the coalitional game model,which could obtain the optimal coalition structure satisfying the user s requests of virtual machines instance;I designed a merge and split rule of coalitions,which could make the final coalition structure stable.In addition,I also designed a normalized estimated Banzhaf value method based on the contribution capacity of coalitional members to divide the coalitional overall profit among its members.The experimental results show that,under the test condition of different virtual machine request type,this cloud coalition formation algorithm can not only get higher overall coalitional profit,but also can make profit division more fair and the execution efficiency higher.

关 键 词：云资源提供 博弈 联盟结构 利益分割 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]