RIS辅助多MEC服务器的联合任务卸载和资源分配策略  

Joint task offloading and resource allocation strategy for RIS-assisted multiple MEC servers

作　　者：王丹[1] 赵静 王嘉程 Wang Dan;Zhao Jing;Wang Jiacheng(School of Communications&Information Engineering,Chongqing University of Posts&Telecommunications,Chongqing 400065,China)

机构地区：[1]重庆邮电大学通信与信息工程学院,重庆400065

出　　处：《计算机应用研究》2025年第2期566-574,共9页Application Research of Computers

基　　金：重庆市自然科学基金创新发展联合基金资助项目(中国星网)(CST2023NSCQ-LZX0114)。

摘　　要：针对传统集中式计算无法有效应对海量设备产生的庞大数据,且移动边缘计算(MEC)服务器资源受限容易引起资源拥塞以及复杂的通信环境导致用户任务卸载传输受阻甚至中断的问题,提出了一种可重构智能表面(RIS)辅助多MEC服务器的联合任务卸载和资源分配方法。由于最大化系统卸载效用问题是一个混合整数非线性规划问题,难以直接求解,于是整体求解过程采用交替优化法,进行迭代求解。通过联合优化RIS处相移矩阵、MEC服务器端中央处理器(central processing unit,CPU)的计算资源、卸载用户和MEC服务器的关联决策以及用户端上行传输功率,最大化系统卸载效用。具体地,通过最佳相位规划,拟凸优化技术以及凸优化技术分别求解RIS最佳相移、用户发射功率分配以及MEC服务器计算资源分配决策,并设计了一种改进的启发式算法求解用户与MEC服务器的关联决策。仿真结果表明,将RIS和改进的启发式算法结合的方法较传统求解方法中的启发式算法相比,系统的平均卸载效用提升了约22.89%,并且方法比基准方案中采用基于局部搜索的经典求解方法的卸载效用提升了约14.02%。因此,该方法有益于提高用户的通信服务质量。Aiming at the problems that traditional centralized computing cannot effectively deal with the huge data generated by massive devices,and the limited resources of MEC server can easily cause resource congestion,and complex communication environment can lead to user task unloading transmission obstruction or even interruption,this paper proposed a method for RIS aided multi-MEC server joint task unloading and resource allocation method.Since the problem of maximizing system unloading utility is a mixed integer nonlinear programming problem,it is difficult to solve directly,so the overall solution process of this problem adopted alternating optimization method and iterative solution.It maximized the system offloading utility through joint optimization of RIS phase shift matrix,MEC server-side CPU computing resources,associated decision-making and clients of uplink transmission power.Specifically,quasi-convex optimization technique and convex optimization techniques respectively solved the RIS best phase shift,user transmission power allocation and MEC server computing resource allocation decisions through the best phase planning,and designed an improved heuristic algorithm to solve the user associated with MEC server s decisions.The simulation results show that the proposed method combining RIS and the improved heuristic improves the average unloading utility of the system by about 22.89%compared with the traditional heuristic algorithm,and the proposed method improves the unloading utility by about 14.02%compared with the classical solution based on local search in the benchmark scheme.Therefore,the proposed method is beneficial to improve the communication service quality of users.

关 键 词：移动边缘计算 可重构智能表面 任务卸载 资源分配 启发式算法 

分 类 号：TN929.5[电子电信—通信与信息系统]