协作式云资源博弈分配

Cloud resource game allocation based on cooperation

针对用户对云资源的异构性需求和竞争问题,提出了一种协作式资源分配博弈策略。建立了资源分配的协作式博弈模型,定义了协作博弈的用户效用函数和评估函数,证明了在该效用函数下协作博弈存在唯一Nash均衡,并讨论了用户组建联盟对协作特征函数和整体效用的影响。实验结果表明,在该协作博弈策略下,个体用户通过组建联盟集体出价的方式,能够为联盟用户带来更大效用,以收敛方式实现Pareto改进。

For the heterogeneity of user requirements and the competition in clouds, a cooperative resource allocation game strategy was presented. The cooperative game model of resource allocation was established and the utility function and the evaluation function of users were defined. Meanwhile, it was proved that there exists unique Nash equilibrium of cooperative game in utility function, and how users coalition effeeted on the charateristic function and the whole utility was discussed. The experimental results show that in this cooperative game strategy, the individual user in the coalition could gain more utility and a Pareto improvement is implemented with a convergence through building a user coalition with multiple users.