云环境中基于布谷鸟搜索算法的多目标任务调度方案

Multi-objective task scheduling scheme based on cuckoo search algorithm in cloud environment

效率往往是任务调度的首要目标,对于数据中心而言,能耗问题也是十分重要的因素。在布谷鸟搜索(cuckoo search,CS)算法的基础上提出了一种多目标任务调度方案——MOCS,以实现云环境下任务调度效率和能耗的Pareto最优。布谷鸟搜索算法是一种启发式算法,利用Lévy flight(莱维飞行)通常能较快地寻找到全局最优解。利用Cloud Sim云仿真平台将所提方案与采用遗传算法的多目标任务调度方案进行对比,仿真实验证明所提方案优于采用遗传算法的方案。

Effectiveness was always the primary goal of task scheduling, for data centers, power consumption was also very important factor. Based on the cuckoo search algorithm, this paper proposed a multi-objective scheduling scheme-MOCS to a- chieve the Pareto optimization between low power consumption and efficiency of scheduling in cloud environment. Cuckoo search algorithm was a heuristic algorithm and could find global optima quickly. It used the CloudSim platfoim to compare the proposed scheme with the scheme employing genetic algorithms. Simulation results show that the proposed scheme outperforms the scheme employing genetic algorithms.