首页 | 本学科首页   官方微博 | 高级检索  
     

基于遗传算法的多处理器系统任务调度
引用本文:司炯,李东生. 基于遗传算法的多处理器系统任务调度[J]. 微型机与应用, 2011, 30(10): 77-79
作者姓名:司炯  李东生
作者单位:1. 合肥工业大学仪器科学与光电工程学院,安徽合肥,230009
2. 合肥工业大学微电子设计研究所,安徽合肥230009;电子工程学院,安徽合肥230031
基金项目:国家863计划项目(2008AA01Z135)
摘    要:用一种遗传算法的调度策略,以大维度矩阵求逆为实验对象,探索在多核中如何完成任务的均衡分配问题,以达到加速效果.算法利用系统资源的弹性,自动搜寻可以并行的子任务并将其合理地分配到相应计算节点中,提高了多核系统资源调度性能,实现了对用户提交的任务的优化调度,达到了均衡系统各处理器计算负载和提高多核系统的总体性能的目标.

关 键 词:多核处理器  遗传算法  任务调度  矩阵求逆

Multiprocessor computing system task scheduling based on genetic algorithms
Si Jiong , Li Dongsheng. Multiprocessor computing system task scheduling based on genetic algorithms[J]. Microcomputer & its Applications, 2011, 30(10): 77-79
Authors:Si Jiong    Li Dongsheng
Affiliation:Si Jiong1,Li Dongsheng2,3(1.School of Instrument Science and Opto-electronics Engineering,Hefei University of Technology,Hefei 230009,China,2.Institute of VLSI Design,3.Hefei Electronic Engineering Institute,Hefei 230031,China)
Abstract:In this paper,we provide a method of task scheduling which depends upon a kind of genetic algorithm,target at the calculation process big dimension matrix' s inverse,explore the methods of how to achieve the aims of balanced assignment and acceleration.The genetic algorithm makes use of the system resources' elasticity,searches automatically the sub-tasks which are parallel and assigns them into the right calculation nodes.It improved the multi-processor system's scheduling performance,realized task's optim...
Keywords:multi-processor  genetic algorithms  task scheduling  matrix inversion  
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号