基于移动边缘计算环境下的服务缓存和任务调度联合优化算法

Joint Optimization Algorithm of Service Cache and Task Scheduling Based on Mobile Edge Computing Environment

在移动边缘计算中,在资源有限的边缘设备上对服务缓存和任务执行进行合理的决策能够大幅度地提高卸载效率和减少应用程序的处理时延;针对边缘计算环境下服务缓存与任务卸载决策问题,建立网络模型和服务缓存模型,定义关联的边缘设备、协作的边缘设备任务执行时延及远端云数据传输和任务执行的时延,提出一种联合优化算法来求解任务执行时延约束条件下的服务缓存决策最优解;该算法采用粒子群优化Particle Swarm Optimization(PSO),将移动用户的任务按照整数编码,优化任务处理时延适应度和粒子速度更新,缩短任务延迟时间;仿真实验结果表明,联合优化算法取得相比其他策略完成时间更少且能适应大规模任务调度的效果。

Making reasonable decisions for service cache and task execution on edge devices with limited resources in the mobile edge computing can greatly improve the unloading efficiency and reduce processing delay of applications.With defining the task execution delay of associated edge devices,cooperative edge devices and remote cloud,a joint optimization algorithm was proposed by establishing a network model and service cache model to solve the optimal solution of service caching decision under the constraint of task execution delay on the problem of service cache and task unloading decision.The algorithm with PSO used integer encoding of mobile users’tasks,optimization of task processing delay fitness and particle velocity updating to shorten the task delay time.The simulation results showed that the algorithm can obtain less completion time in the large-scale task scheduling than other strategies.