一种改进的基于复制的异构多核任务调度算法

An Improved Duplication Based Heterogeneous Multi-core Scheduling Algorithm

基于复制的异构多核任务调度算法普遍存在复制的冗余调度过多、处理器利用率不高,产生大量能耗等问题。目前已有一些算法针对该问题对冗余调度进行优化,然而这些优化算法存在检测冗余时机晚、优化空间小、时间复杂度高等问题。针对这些问题提出一种改进的基于复制的异构多核任务调度算法,采用二路复制策略并结合冗余处理机制,旨在最小化调度长度,同时减少冗余的复制任务数量。实验结果表明,该算法在减少冗余任务量和最小化任务调度长度两个方面具有较好的性能。

Duplication based heterogeneous multi -core task scheduling algorithms generally suffers redundant scheduling, low processor utilization, and high energy consumption. At present, some algorithms have been proposed to optimize the redundant scheduling problem. However, the detection of the redundancy by these algorithms is late with high time complexity, leaving little space for optimization, and. An improved duplication based heterogeneous multi - core task scheduling algorithm is proposed by combining the binary duplicating policy with the redundant processing mechanism to minimize the scheduling length while reducing the number of redundant replication tasks. Experimental results show that the proposed algorithm has better performance in aspects of reducing redundant tasks and minimizing the length of the task scheduling.