基于蚁群算法的云任务调度研究  

Research of Cloud Task Scheduling Based on Ant Colony Algorithm

作　　者：王云松[1] 孙佳林[1] 龚跃[1] 

机构地区：[1]长春理工大学计算机科学与技术学院,长春130022

出　　处：《长春理工大学学报（自然科学版）》2017年第1期133-136,共4页Journal of Changchun University of Science and Technology(Natural Science Edition)

摘　　要：针对云计算中任务分配算法效率不高的问题,提出了一种改进的蚁群算法来解决云计算中的任务分配问题。首先假定要分配的任务为蚂蚁的起点,执行任务的虚拟机为蚂蚁的终点,任务分配的过程就是蚂蚁从起点走到终点的过程。然后随机选择一个任务作为蚂蚁的起点,用改进的蚁群算法计算后把任务分配给相应的虚拟机,直到所有任务都分配完成。最后当所有蚂蚁都把任务分配完成后,选择代价最小的路径作为本次任务分配的方案。通过使用cloudsim仿真器进行仿真实验,证明了蚁群算法能够有效的解决云计算中任务分配的问题。In order to solve the problem of low efficiency of task allocation, an improved ant colony algorithm was pro-posed to solve the problem of task allocation in cloud computing.Firstly, It was assumed that the task was the starting point of the ants, the virtual machine to perform the task was the end of the ants, the process of task allocation was the process of ants from the beginning to the end.Secondly, one task was selected as the starting point of the ants ran- domly.The improved ant colony algorithm was used to assign the task to the corresponding virtual machine.Until the end of the task assignment was completed.Finally, the path of the minimum cost was chosen as the solution of this task when all the ants were assigned to the tasks.By using the cloudsim simulator, it is proved that ant colony algo- rithm can effectively solve the problem of task allocation in cloud computing.

关 键 词：蚁群算法 云计算 任务分配 cloudsim 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]