摘要
针对受限云资源的全局优化调配问题,研究提出了相应的调配模式、模型和算法;首先针对云资源调配问题的新特征,提出了整体资源受限条件下,以整体效用最佳为决策目标、多任务多资源同时调配的全局优化调配模式;接着引入了多Agent系统理论的基本原理,研究建立了受限云资源的全局优化调配模型;继而设计了基于矩阵编码的遗传算法来求解该问题模型;仿真实验结果表明,该算法能有效、高效地求解受限云资源的全局优化调配问题。
To tackle the problem of global optimal--allocation of limited cloud resources （GOALCR） within cloud computing, the corresponding resource allocation mode, mathematical model and algorithm were studied. In consideration of the new characteristics of GOALCR, a global optimization mode characterized by multi--task multi--resource allocation was proposed under the objective of optimal overall utility and the assumption of limited overall resources. Then, the problem of GOALCR was formulated in the lights of the basic concepts of multi-- agent system, followed by the design of the matrix coded genetic algorithm to solve the problem. Ultimately, the experimental results demonstrated the proposed algorithm was both efficient and effective.
出处
《计算机测量与控制》
北大核心
2013年第2期523-525,共3页
Computer Measurement &Control
关键词
云计算
整体资源受限
全局优化调配
遗传算法
cloud computing
limited overall resources
global optimal--allocation
genetic algorithm
