论文部分内容阅读
针对可迁移依赖任务的重调度问题,提出了基于约简DAG可迁移任务图的重调度模型,并基于免疫遗传算法实现了以提高应用性能为目标的求解算法.实验表明,与经典的动态调度算法Max-Min和基于启发式的AHEFT静态算法相比较,由于调度目标的一致性,初始调度的性能在重调度过程中被较好地保持,并且由于任务迁移的支持和遗传算法在全局优化上的性能优势,应用性能得到较大提升;又由于任务图的约减过程和免疫因子对算法收敛的作用,提出的IGA算法效率得到显著改善,使资源动态性和异构性的适应能力得到进一步增强.
Aiming at the re-scheduling problem of relocatable dependency tasks, a re-scheduling model based on the reducible task graph of DAG is proposed and an algorithm based on immune genetic algorithm is proposed to improve the performance of the algorithm.Experiments show that, compared with the classical dynamic scheduling Compared with the heuristic AHEFT static algorithm, Max-Min algorithm is better than the heuristic AHEFT algorithm because of the consistency of scheduling objectives. The performance of initial scheduling is better maintained in the process of re-scheduling. Because of the support of task migration and the global optimization of genetic algorithm The performance of IGA algorithm has been greatly improved; and due to the reduction process of the task graph and the effect of immune factors on the convergence of the algorithm, the proposed IGA algorithm has been significantly improved in efficiency, which further improves the adaptability of resource dynamics and heterogeneity Enhanced.