基于混合并行布谷鸟搜索的作业调度算法

Job scheduling algorithm based on hybrid parallel cuckoo search

针对布谷鸟搜索算法在解决作业调度问题时存在收敛速度慢等问题,提出一种基于混合并行布谷鸟搜索的作业调度算法。在传统布谷鸟搜索基础上,结合多核集群的拓扑结构特点,将消息传递编程模型MPI与共享存储编程模型OpenMP结合,采用基于作业优先级的编码方式将布谷鸟搜索算法中的鸟巢位置转换为作业调度序列进行求解。仿真结果表明,与传统的串行布谷鸟搜索算法相比,使用该方法求解作业调度问题的收敛速度有明显加快。

Cuckoo search algorithm has been proved with a slow convergence speed in solving job scheduling. Based on this, a job scheduling algorithm based on hybrid parallel cuckoo search algorithm was proposed. MPI was combined with OpenMP, which made full use of the structural features of multi-core cluster. The nest position was translated into a job scheduling sequence using the coding method based on job priority. Simulation results show that the proposed algorithm has higher convergence speed compared with serial cuckoo search algorithm.