刘文帅
1，李斌
1,2
（1. 南京信息工程大学，中国 南京 210044；2. 网络与交换技术国家重点实验室（北京邮电大学），中国 北京 100876）
摘要：车载边缘计算（VEC）是融合车联网与移动边缘计算的一种全新范式。针对障碍物遮挡对车联网中路边单元（RSU）服务性能的影响，提出一种智能超表面（RIS）辅助的VEC部分任务卸载方案。首先，综合分析车辆移动性对系统的影响；其次，联合优化发射功率、卸载比例和时段划分，旨在建立一个车辆最小平均速率最大化问题；最后，采用一种基于PPO的深度强化学习（DRL）方法求解该优化问题。仿真结果表明，相比随机时段划分策略，所提算法的最小平均速率和卸载比例分别提升了61.9%和46.8%。
关键词：智能反射面；车载边缘计算；深度强化学习；任务卸载
Reconfigurable Intelligent Surface-Enabled Vehicular Edge Computing
LIU Wenshuai1, LI Bin1,2
(1. Nanjing University of Information Science and Technology, Nanjing 210044, China; 2. State Key Laboratory of Networking and Switching Technology (Beijing University of Posts and Telecommunications), Beijing 100876, China)
Abstract: Vehicular edge computing (VEC) is a new paradigm of integrating the Internet of vehicles and mobile edge computing. To compensate the influence of obstacle occlusion on the service performance of road side units (RSUs) in the internet of vehicles, a partial offloading scheme based on Reconfigurable Intelligent Surface (RIS) is proposed. First, the mobility of vehicles is considered. Then, the minimum average-rate maximization of the vehicles is formulated. Finally, the proposed optimization problem is solved by PPO driven deep reinforcement learning (DRL) method. Simulation results show that the proposed algorithm improves the minimum average rate and offloading ratio by 61.9% and 46.8%, compared with the random time division strategy.
Keywords: reconfigurable intelligent surface; vehicular edge computing; deep reinforcement learning; task offloading
.png)
.png)
.png)
