异构分布式系统的可分任务调度算法  

New Genetic Algorithm for Divisible Load Scheduling in Heterogeneous Distributed Systems

作　　者：王晓萍[1] 孟坤[2] 

机构地区：[1]西安文理学院软件学院,西安710068 [2]西安电子科技大学计算机学院,西安710071

出　　处：《小型微型计算机系统》2015年第4期797-800,共4页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61402350)资助;西安市科技计划创新基金文理专项项目(CXY1352WL07)资助

摘　　要：针对异构分布式系统下带数据结果收集的可分任务调度问题,通过设计一种新的遗传算法,包括编码与解码方案、交叉和变异操作,有效地解决了3个难点:(1)计算所需的处理机数目及最优的处理机选择;(2)给出最优的任务分配顺序;(3)计算每个处理机需要处理的任务大小.为了验证新算法的有效性,本文进行了一系列对比实验.结果表明任务分配序列对完成时间具有显著的影响,且新算法能够显著降低任务的完成时间.The problem of scheduling divisible loads with results collection in network based heterogeneous distributed systems is ad- dressed in this paper by designing a new genetic algorithm,inclusive of encoding and decoding methods,crossover and mutation opera- tions. The proposed algorithm effectively tackles the following three issues： （ 1 } how many and which processors are required in com- putation; （ 2 ） in which order the load fractions are distributed to processors ; （ 3 ） how much the load fraction should be distributed to each processor. To show the effectiveness of the algorithmla set of comparative experiments are conducted. Experimental results show that the distribution sequence has a significant influence on the processing time and the proposed algorithm can effectively reduce it.

关 键 词：可分任务 分发顺序 完成时间 遗传算法 异构分布式系统 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]