云环境下改进SOS的多目标任务调度算法

Multi objective task scheduling algorithm based on improved SOS in cloud environment

为有效获取云计算中多目标任务调度求解算法的全局最优解,提出一种云环境下基于改进期望服务质量(Qo S)的多目标任务调度算法。设计多目标任务调度框架,提出相应的目标函数与约束条件。利用准反射学习构建初始种群以改进共生生物搜索(SOS)算法,加入自适应变异率以提高全局搜索能力。通过设定坐标进行任务分配,利用改进后SOS算法实现多目标任务优化调度。云计算仿真结果表明,所提算法相比于其它算法,有效改善了能源利用率、能耗和时间成本,具有较好的Qo S传输性能。

To effectively obtain the global optimal solution of the multi-objective task scheduling algorithm in cloud computing,a multi-objective task scheduling algorithm based on improved expected quality of service(QoS)in cloud environment was proposed.A multi-objective task scheduling framework was designed,and the corresponding objective function and constraint conditions were proposed.The quasi-reflection learning was used to construct the initial population to improve the Symbiosis Search(SOS)algorithm,and the adaptive mutation rate was added to improve the global search ability.The task was allocated by setting the coordinates,and the optimized SOS algorithm was used to realize the optimal scheduling of multi-objective tasks.Cloud computing simulation experiment results show that compared with other algorithms,the proposed algorithm effectively improves energy utilization,reduce energy consumption and time cost,and has better QoS transmission performance.