摘要
为了实现云计算环境下的集群资源调度和实现资源负载平衡,提出了一种基于克隆选择算法的云计算集群资源调度方法。首先,定义了以最小化执行时间跨度和负载均衡因子为目标的云计算资源调度模型。在此基础上提出了一种采用克隆选择算法对云计算环境下集群资源进行调度的方法,对抗体编码方式,抗体与抗体之间以及抗体与抗原之间的亲和度函数、免疫克隆算子、退火交叉算子以及高斯变异算子均进行了设计。并定义了采用改进的克隆选择算法进行集群资源调度的具体算法。仿真实验表明:方法能获得最优的资源调度方案,且与其它方法相比,具有较少的执行时间跨度和负载均衡因子,具有较大的优越性。
In order to realize the cluster resource scheduling in cloud computing, a cluster resource scheduling method was proposed based on clone selection algorism. Firstly, the scheduling model was defined based on the goal by minimizing execution time span and resource load balance factor was designed, and then the cluster re- source scheduling method based on clone selection algorism was introduced. The code method, the affinity func- tions between antibodies and antibodies and the ones between antibodies and antigens, immune clone operator, an- nealing crossover operator and Gaussian mutation operator were designed. Finally, the algorism based on clone selection algorism for cluster resource scheduling was given. The simulation result shows the method can get the best task scheduling program, and compared with the other methods, having the short execution time span and load balance factor with big priority.
出处
《科学技术与工程》
北大核心
2013年第13期3642-3646,共5页
Science Technology and Engineering
基金
国家自然基金项目(60211227)
江苏省教育科学"十二五"规划2011年度课题(D/2011/03/006)资助
关键词
资源调度
云计算
克隆选择算法
负载均衡
resource scheduling cloud computing clone selective algorism load balance
