基于预测及蚁群算法的云计算资源调度策略  

Cloud Computing Resource Scheduling Strategy Based on Prediction and ACO Algorithm

作　　者：周文俊[1] 曹健[1] 

机构地区：[1]上海交通大学电子信息与电气工程学院,上海200240

出　　处：《计算机仿真》2012年第9期239-242,246,共5页Computer Simulation

基　　金：国家自然科学基金(61073021);上海市科委项目(10511501503;10DZ1200200;11511500102)

摘　　要：研究云计算资源调度问题,针对目前静态的网格资源调度算法只考虑任务完成时间最小化,导致了不能满足动态的云计算资源调度要求。为了适应云计算的动态性和实时性,解决云计算资源调度问题,降低数据中心用电量,提出一种基于预测及蚁群算法的云计算资源调度策略。当数据中心利用率较低时运行改进蚁群算法来合理调度虚拟机至宿主机,通过动态趋势预测算法预测数据中心负载来智能开关宿主机。仿真结果表明,采用预测及蚁群算法进行的云计算资源调度策略,保证了云计算的实时性,并有效减少数据中心用电量。Cloud computing resource scheduling was studied. The current static grid resource scheduling algo-rithms consider only the minimization of the makespan, so they can not meet the demands of cloud computing resource scheduling. In order to adapt to the dynamic and real-time nature and solve the issue of cloud computing resource scheduling and decrease the power consumption of datacenter, we proposed a cloud computing resource scheduling al- gorithm based on prediction and ACO algorithm. When the utility of datacenter is low, the improved ACO algorithm is executed to assign VMs to hosts. Dynamic tendency prediction strategy was used to predict the load of datacenter and turn on/off hosts. The results of simulation show that in the case of running VMs normally, cloud computing resource scheduling strategy based on prediction and ACO algorithm can reduce the power consumption of datacenter effective-ly.

关 键 词：云计算 资源调度 预测算法 蚁群算法 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]