摘要
针对网格环境下用户难以获得资源竞价所需的信息而导致的决策风险,将不完全信息资源竞价转化成完全信息下的重复博弈问题。分析了该博弈均衡解的存在性及求解过程,给出了相应的竞价算法,讨论了对用户低价联盟的抑制方法。仿真实验表明用户通过各阶段资源预配置的信息调整竞价策略,资源配置可逐步逼近均衡解,实现网格资源的优化配置。
Because the grid users are impossible to collect all important information in real grid environments, they may make the wrong decision in real grid environments. To address this issue, this problem is formulated as a two - player repeated game. The existence and the derivation of equilibrium are proposed, the gird user and broker bidding algorithms are presented, and the outcome of hostile takeover is also discussed. The simulate results show that grid user can adjust bidding strategy constantly on each stage of repeated game to achieve equilibrium, which leads to an optimal allocation of grid resource.
出处
《智能计算机与应用》
2014年第1期6-9,共4页
Intelligent Computer and Applications
关键词
不完全信息
重复博弈
网格计算
资源配置
Incomplete Information
Repeated Game
Grid Computing
Resource Allocation
