移动边缘计算中的多用户多边协同卸载策略

Multi-user and Multilateral Cooperative Offloading Strategies in Mobile Edge Computing

在多用户多边缘服务器的边缘计算场景下,用户设备在任务卸载时,由于卸载目标的不确定性常导致移动边缘服务器负载不均衡,出现资源紧张或资源空闲浪费问题,甚至导致卸载决策失效。针对此问题,提出了面向时延和能耗感知的启发式任务多边协同卸载(HTMC)算法,并在算法中提出了一种基于粒子近期历史位置的位置更新策略(PRPPUS)。该算法以最小化UE能耗和任务执行时延的加权和为目标,根据时延和能耗感知自适应地选择任务卸载策略,达到最小化UE开销目标。仿真实验结果证明,与采用粒子群算法和遗传算法的卸载策略比较,所提HTMC算法的性能更高、更平稳,算法运行开销较小,优化效果更突出。

In the multi-user and multi-edge server edge computing scenario,the uncertainty of the offload target of User Equipment(UE)often leads to an unbalanced load on Mobile Edge Computing Server(MECS),resource constraint or idle waste,and even leads to the offload decision fails.To address this problem,a delay and energy-aware Heuristic Task Multilateral Co-Offloading(HTMC)algorithm is proposed,and a Particle Recent Past-based Position Updating Strategy(PRPPUS)is proposed in the algorithm.The algorithm aims to minimize the weighted sum of UE energy consumption and task execution delay,and adaptively selects the task offloading strategy according to the delay and energy awareness to minimize UE overhead.The simulation experimental results prove that the proposed HTMC algorithm has higher and smoother performance,smaller algorithm running overhead,and more outstanding optimization effects compared with the offloading strategies using the Particle Swarm Optimization algorithm and Genetic Algorithm.