摘要
本文讨论一种启发式任务分配方法，称之为改进的ｌｉｓｔ分配方法，它适用于分配一组具有先后关系和通信延迟的任务集到同构型分布式计算机系统上．文中描述了此分配方法的原理和算法，给出相应的仿真流程图，并对具有不同拓扑结构、任务运行时间和通信时间满足多种概率分布的任务集进行了分析和仿真．结果表明，当处理器个数小于任务集的并行度、任务粒度大于５时，任务分配效率大于８０％．
This paper describes a heuristic algorithm of task allocation,called the improved list scheduling. It is used to allocate tasks with precedence order and the communication delay onto a homogeneous distributed computer system. The algorithm is analysed and simulated for various tasks with different topological structures and distributions of task execution time and communication time. The result shows that task allocation efficiency is greater than 80%,if the number of processors is fewer than the parallelism of tasks and the task granularity is greater than 5.
出处
《计算机学报》
EI
CSCD
北大核心
1994年第2期112-119,共8页
Chinese Journal of Computers
关键词
任务分配
启发式算法
分布式计算机
Task allocation,precedence tasks,heuristic algorithm,homogeneous distributed computer
