车联网中基于多智能体强化学习的边缘服务器选址策略

Edge server deployment strategy based on multi-agent reinforcement learning in the internet of vehicles

为解决车联网环境下边缘服务器选址难的问题,提出一种基于多智能体强化学习的边缘服务器部署策略(记为CKM-MAPPO),重点优化边缘服务器间的负载均衡,同时最小化边缘服务器的时延和能耗。首先,使用Canopy和K-means算法确定边缘服务器部署的数量和初始位置;然后,基于多智能体强化学习算法确定边缘服务器的最优部署位置;最后,通过一系列实验评估所提出算法的准确性和有效性。研究结果表明:与基准算法相比,本文提出的方法的负载均衡度提升了26.5%,时延和能耗分别降低了12.4%和17.9%。

To solve the hard problem of edge server deployment in internet of vehicle environments,an edge server deployment strategy based on multi-agent reinforcement learning(CKM-MAPPO)was proposed.It focuses on optimizing the load balancing among edge servers and minimizing edge servers'delay and energy consumption.Firstly,the Canopy and K-means algorithms were used to determine the number and initial location of edge server deployment.Then,the multi-agent reinforcement learning algorithm was leveraged to determine the optimal deployment location of the edge server.Finally,the accuracy and effectiveness of the proposed algorithm were evaluated through a series of experiments.The results show that compared with the benchmark algorithm,the proposed method improves load balancing by 26.5%,and the time delay and energy consumption are reduced by 12.4%and 17.9%,respectively.