多机异构相关任务集的调度优化研究

OPTIMAL SCHEDULING OF MIXED STRUCTURE TASK SET IN MULTIPROCESSORS

本文首先介绍了异构相关任务集的概念,在实现子任务优化调度的前提下,基于相同处理机(identical processors)并根据多目标优化策略,提出了AOSA调度算法,构成多机相关任务集的优化调度方法。实验表明:(1)AOSA方法优于随机非AOSA方法对异构任务集的调度;(2)理论准则与实验结果一致。

Preserving the optimal scheduling of sub-task set and using multirobject optimization strategies,this paper proposes the advanced optimal schedule assembly (AOSA) algorithm to get the optimal scheduling of mixed structure task set in identical multiprocessors.The experimental result confirms with the theoretical one and shows that AOSA algorithm is better than any random non-AOSA methods.