摘要
给出一种对异构计算系统(HCS)进行任务映射与调度的遗传算法——SMT-GA算法.首先对HCS任务调度问题作出形式描述,然后分别介绍SMT-GA算法的总体框架,染色体设计,从染色体获取调度方案的方法,染色体适合度函数设计,交叉与变异遗传操作算子设计等.最后给出算法实现情况及实验结果分析与结论等.
The SMT GA algorithm, which is a genetic algorithm for task mapping and scheduling in Heterogeneous Computing Systems (HCS), is introduced. First the task scheduling problem for HCS is described. Then the structure of the algorithm, chromosome design, the method to generate scheduling solution from a chromosome, the design of fitness function, the crossover and mutation operator design, are discussed respectively. In the end, the algorithm realization, experiment result analysis and the conclusions are presented.
出处
《小型微型计算机系统》
CSCD
北大核心
1999年第4期241-245,共5页
Journal of Chinese Computer Systems
基金
国家自然科学基金
关键词
异构计算系统
任务调度
SMT-GA
算法
Heterogeneous computing Task scheduling Genetic algorithm Multiprocessor scheduling Network parallel computing