能量收集技术驱动的移动边缘计算卸载策略  

COMPUTATION OFFLOADING STRATEGY DRIVEN BY ENERGY HARVESTING TECHNOLOGY IN MOBILE EDGE COMPUTING

作　　者：程海涛 柯鹏[1] Cheng Haitao;Ke Peng(College of Computer Science and Technology,Wuhan University of Science and Technology,Wuhan 430065,Hubei,China)

机构地区：[1]武汉科技大学计算机科学与技术学院,湖北武汉430065

出　　处：《计算机应用与软件》2024年第12期77-85,160,共10页Computer Applications and Software

摘　　要：针对配有能量收集设备的多用户MEC系统中的任务卸载问题,提出一种在线计算卸载策略。通过Lyapunov优化将原始的随机优化问题进行转化,从而计算出移动设备在每个时隙的最佳CPU频率和发射功率,再使用结合贪心策略的模拟退火算法找到任务的最佳执行位置。实验结果表明,与任务完全在本地执行和贪心选取执行位置两种策略相比,所提算法将系统执行成本分别降低了44.1%和20.3%。与传统的LODCO算法和SDTO算法相比,任务卸载率分别提高7.3百分点和4.1百分点,保证了用户的QoE。Aimed at the task offloading problem in multi-user MEC system equipped with energy harvesting equipment,an online computation offloading strategy is proposed.Lyapunov optimization was utilized to transform the original stochastic optimization problem,so as to calculate the best CPU frequency and transmit power of the mobile devices in each time slot.The simulated annealing algorithm combined with the greedy strategy was used to find the best execution position of the tasks.The experimental results show that compared with the two strategies of task execution completely locally and greedy selection of execution position,the proposed algorithm reduces the system execution cost by 44.1%and 20.3%respectively;compared with the traditional LODCO algorithm and SDTO algorithm,task offloading rate increase by 7.3 and 4.1 percentage points respectively,ensuring the user s QoE.

关 键 词：移动边缘计算 能量收集 计算卸载 Lyapunov优化 模拟退火算法 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]