移动边缘计算中支持能量收集的计算卸载策略  

Computing offloading strategy supporting energy collection in mobile edge computing

作　　者：蒋欣秀 杨俊东[1] 杨志军[1] 李波[1] 丁洪伟[1] JIANG Xinxiu;YANG Jundong;YANG Zhijun;LI Bo;DING Hongwei(Yunnan University,Kunming 650500,China)

机构地区：[1]云南大学,云南昆明650500

出　　处：《现代电子技术》2022年第1期17-23,共7页Modern Electronics Technique

基　　金：国家自然科学基金资助项目:融合式多址通信网络理论与控制协议研究(61461053)。

摘　　要：移动边缘计算(MEC)作为一种新的范式受到了各界的关注,但时延敏感、能耗巨大,MEC的局限性也日益凸显,为了解决上述问题,提出一种可由能量收集技术获取能源进行供能的边缘计算系统模型。首先,将时延、能耗和任务丢弃率作为指标,建立执行代价模型;然后引入能量收集技术模拟用户能量收集过程;最后针对模型时域耦合问题,利用带扰动Lyapunov理论将问题转化为逐时隙定性问题,并基于改进的灰狼算法对CPU频率和发射功率进行迭代以获取最小的任务执行代价。实验结果表明,改进的灰狼算法与其他算法相比,其任务执行代价更小,用户体验更好。As a new paradigm,the mobile edge computing(MEC)has attracted the attention of all walks of life. However,the limitations of MEC has been becoming increasingly prominent with time-delay sensitivity and huge energy consumption. In view of the above,an edge computing system model which can obtain energy by energy collection technology and supply the energy is proposed. The execution cost model is established by taking the delay,energy consumption and task lose rate as indicators,and then the energy collection technology is introduced to simulate the energy collection process of users. The timedomain coupling problem of the model is transformed into a time slot qualitative problem by the Lyapunov theory. The CPU frequency and the transmitted power are iterated based on the improved gray wolf optimization(IGWO)algorithm to obtain the lowest task execution cost. The experimental results show that,in comparison with the other algorithms,the IGWO algorithm has the lower task execution cost and the better user experience.

关 键 词：计算卸载 能量收集 移动边缘计算 改进灰狼算法 功率迭代 执行代价模型 

分 类 号：TN911.1-34[电子电信—通信与信息系统] TP399[电子电信—信息与通信工程]