云计算环境下用户任务调度效率优化仿真

发布时间：2018-03-24 04:08

本文选题：云计算环境　切入点：复杂任务　出处：《计算机仿真》2017年04期

【摘要】：为提高用户任务处理的整体速度,需对云计算环境下用户任务调度效率进行优化。进行用户任务调度效率优化时,应对全部虚拟机执行复杂任务调度所用时间进行确定,最终确定任务调度执行总成本目标函数,但是传统方法以复杂任务与资源节点属性的配合程度为准则,对复杂任务进行聚类求解完成优化,但是不能确定任务调度执行总成本目标函数,导致任务调度效率优化时总成本较高,效果差的问题。提出一种基于蝙蝠算法的云计算环境下用户任务调度效率优化方法,首先获取云计算环境下虚拟机复杂任务资源集合,组建复杂任务时间矩阵和资源调度矩阵,计算出全部虚拟机执行复杂任务调度所用时间,结合虚拟机定价模式给出每个虚拟机时间单位成本,确定任务调度执行总成本目标函数,引人蝙蝠理论将蝙蝠个体视为目标函数优化问题求解空间中的搜索体,将复杂任务分割成若干个子任务,定义子任务的编号规则,并定义了适应值函数,利用优化问题的适应值函数评价蝙蝠个体位置的优劣,依据评价的结果对用户任务调度效率进行优化。仿真结果表明,改进方法减少了复杂任务调度的总完成时间,提高了计算资源的利用率。
[Abstract]:In order to improve the overall speed of user task processing, it is necessary to optimize the efficiency of user task scheduling in cloud computing environment. Finally, the target function of total cost of task scheduling is determined, but the traditional method is based on the coordination degree between complex tasks and resource node attributes, and the complex tasks are solved by clustering to complete the optimization. However, the objective function of total cost of task scheduling can not be determined, which leads to the problem of high total cost and poor effect in task scheduling efficiency optimization. A bat algorithm-based optimization method for user task scheduling efficiency in cloud computing environment is proposed. Firstly, the complex task resource set of virtual machine in cloud computing environment is obtained, the complex task time matrix and resource scheduling matrix are constructed, and the time used by all virtual machines to execute complex task scheduling is calculated. Combined with the pricing model of virtual machine, the unit cost of each virtual machine is given, and the objective function of total cost of task scheduling is determined. The bat theory treats bat individual as a search object in the solving space of objective function optimization problem. The complex task is divided into several subtasks, the numbering rules of the subtasks are defined, and the fitness function is defined. The fitness function of the optimization problem is used to evaluate the position of the bat individual. According to the evaluation results, the efficiency of user task scheduling is optimized. The simulation results show that the improved method reduces the total completion time of complex task scheduling and improves the utilization of computing resources.
【作者单位】：上海电机学院;
【分类号】：TP18;TP393.09

【相似文献】