资源获胜率保证的网格任务调度算法  

Victorious-directed tasks scheduling algorithm for grid computing based on game theory

作　　者：姚磊[1] 戴冠中[1] 王丹丹[2] 张慧翔[1] 

机构地区：[1]西北工业大学自动化学院,西安710072 [2]广东工业大学应用数学学院,广州510006

出　　处：《计算机工程与应用》2009年第30期84-87,共4页Computer Engineering and Applications

基　　金：国防基础科研项目

摘　　要：为保证多资源竞争下单个资源的获胜率,建立一个资源富裕型的网格计算环境模型,基于此模型提出了一个基于博弈论的资源竞争获胜率保证算法(GVP)。GVP算法通过对已知信息分析,预测对手在下一次博弈中的行动,并在此基础上确定自己的行动,使资源累计的获胜率维持在一个可接受的范围。通过实验比较了使用不同竞争策略资源的获胜率。对实验结果进行了深入分析,讨论了两资源博弈的纳什均衡解。In this paper,a model of a resource-rich environment of the grid computing is established at first,and then a new type of grid resource competition algorithm based on this environment is put forward,the Guarantee of Victorious Probability algorithm（GVP）.This algorithm can predict the action of an adversary through known information,and determine itself action based on the forecast.The experiments show that the resources using GVP can be close to their expectations of victorious probability com- pared with the other resources using the other algorithm.A more in-depth analysis of the experimental results is made,and the nash equilibrium of two-resource game is also discussed.

关 键 词：网格计算 博弈论 纳什均衡 获胜率保证 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]