摘要
引入微观经济学与遗传工程知识,兼顾时限与成本,设计了一种网格中的作业分配方法.首先基于拍卖模型确定资源购买者和资源提供者之间的资源交易价格,然后使用遗传算法寻找作业分配最优方案.仿真结果表明,该方法是可行和有效的,不仅效用较高,而且作业对资源的分配较均衡,优于PRIMAL方法.
In this paper,with introduction of microeconomics and genetic engineering knowledge,a job assignment method for grid computing is proposed,considering both time limit and cost simultaneously.It determines resource trading price between resource buyer and resource provider based on auction model,and then finds the optimal job assignment solution based on genetic algorithm.Simulation results have shown that the proposed method is both feasible and effective with higher utility and much balanced job assignment to resource compared with PRIMAL algorithm.
出处
《华中科技大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2006年第z1期9-12,共4页
Journal of Huazhong University of Science and Technology(Natural Science Edition)
基金
新世纪优秀人才支持计划资助项目
国家自然科学基金资助项目(60473089)
中国教育科研网格China-Grid项目
国家发改委CNGI示范工程资助项目(CNGI-04-15-7A
CNGI-04-13-2T和CNGI-04-6-2T)
关键词
网格
作业分配
资源定价
拍卖
遗传算法
grid
job assignment
resources pricing
auction
genetic algorithm