In this paper, we investigate a MEC relay-assisted system with multiple relay nodes (RNs) and multiple remote servers (RSs), where both the selections of best RN and RS are considered. In order to explore the diversity of both RNs and RSs, we propose a sequential relay-remote selection and offloading strategy by taking the local computing, relay computing and remote computing into consideration. The sequential relay-remote selection strategy specifies when to stop server discovery and carry out computation offloading. The offloading strategy tells what are the optimal portions of task executed locally, at RN and at RS. Under such framework, we seek to minimize the total energy consumption coming from server detection, data transmission and data execution. We formulate this problem as a stochastic sequential decision-making problem and dynamic programming is applied to obtain the optimal strategy. The performance of our proposed strategy is evaluated using simulation results. It is found that the energy consumption can be reduced by jointly considering the design of sequential relay-remote selection and offloading ratio optimization.

在本文中，我们研究了具有多个中继节点 (RN) 和多个远程服务器 (RS) 的 MEC 中继辅助系统，其中考虑了最佳 RN 和 RS 的选择。为了探索RN和RS的多样性，我们考虑了本地计算、中继计算和远程计算，提出了一种顺序中继-远程选择和卸载策略。顺序中继远程选择策略指定何时停止服务器发现并执行计算卸载。卸载策略告诉在 RN 和 RS 本地执行的任务的最佳部分是什么。在这样的框架下，我们寻求最小化来自服务器检测、数据传输和数据执行的总能耗。我们将此问题表述为随机顺序决策问题，并应用动态规划来获得最佳策略。使用仿真结果评估我们提出的策略的性能。发现通过联合考虑顺序继电器-远程选择和卸载比优化的设计可以降低能耗。