论文部分内容阅读
多处理器调度问题是影响系统性能的关键问题,基于任务复制的调度算法是解决多处理器调度问题较为有效的方法.文中分析了几个典型的基于任务复制算法,提出了基于动态关键任务(DCT)的多处理器任务分配算法.DCT算法以克服贪心算法不足为要点,调度过程中动态计算任务时间参数,准确确定处理器的关键任务,以关键任务为核心优化调度,逐步改善调度结果,最终取得最优的调度结果.分析和实验证明,DCT算法优于现有其它同类算法.
Multi-processor scheduling problem is the key issue that affects system performance, and scheduling algorithm based on task replication is a more effective method to solve the multi-processor scheduling problem.In this paper, several typical task-based replication algorithms are analyzed, and based on the dynamic mission-critical DCT) algorithm.DCT algorithm to overcome the lack of greedy algorithm as the key point, the process of dynamic scheduling task time parameters to accurately determine the key tasks of the processor, the key tasks as the core of optimal scheduling and gradually improve the scheduling results, Finally, the optimal scheduling result is obtained.Analysis and experiments show that the DCT algorithm is superior to other existing algorithms.