云计算环境下低消耗的任务调度方法仿真  

Simulation of Task Scheduling Method with Low Consumption under Cloud Computing Environment

作　　者：张敏[1] 陈基雄[1] 成亚玲[1] 

机构地区：[1]华中科技大学计算机科学与技术学院,湖北武汉430074

出　　处：《计算机仿真》2015年第10期443-445,451,共4页Computer Simulation

摘　　要：在云计算环境下低消耗的任务调度时,由于云计算环境下低消耗的任务调度具有较高的复杂性和通用性,任务之间存在的多种特征,造成任务调度时受到多种制约条件。传统任务调度方法受到这种特征条件约束,导致任务调度效率降低,无法有效实现任务调度,提出基于改进免疫进化算法的任务调度方法,分析了一般免疫进化方法,将PSO算法看作是一个算子引入免疫进化算法中,对免疫算法进行改进,给出任务调度优化模型。在解空间中随机形成初始抗体,对种群规模、变异概率及最大迭代代数进行初始化处理。求出抗体群中各抗体的亲和度并进行排列,将某抗体进行变异,形成新抗体群,重新对亲和度进行计算和排列,从新抗体群中选择最佳抗体。通过PSO算子对新抗体进行处理,获取一组改进后的抗体,对亲和度较大的抗体进行统计。仿真结果表明,所提方法使云计算调度具有很高的效率。A task scheduling method based on improved immune evolutionary algorithm is proposed in the paper. The general immune evolutionary method was analyzed, the PSO algorithm was taken as an operator and to be introduced into the immune evolutionary algorithm. The immune algorithm was improved, and the task scheduling optimization model was provided. The initial antibody was randomly formed in the solution space, and the population size, mutation probability and the number of generation of maximum iteration were initialized. The affinity of each antibody in antibody population was solved and arranged. Certain antibody was mutated, to form new antibody population. The affinity was calculated and arranged again, and the best antibody was selected from the new antibody population. By using PSO operator, the new antibodies were processed, to acquire a set of improved antibody and make statistics of antibodies with larger affinity. The simulation results show that the proposed method used in cloud computing scheduling has high efficiency.

关 键 词：云计算 低消耗 任务调度 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]