重庆邮电大学学报（自然科学版）2026，Vol.38Issue(3)：391-400,10.DOI:10.3979/j.issn.1673-825X.202505110113
基于深度强化学习的基站协作与服务缓存研究
Research on base station collaboration and service caching based on DRL
摘要
Abstract
Edge caching technology,which stores content in proximity to end users,can effectively reduce transmission redundancy and service latency,making it one of the key enabling technologies for mobile edge networks.Given the limited cache capacity of individual base stations,cooperative edge caching can significantly improve storage utilization through resource sharing among base stations.To address the joint optimization problem of base station cooperation and service caching in mobile edge networks,this study proposes an efficient dynamic decision-making algorithm that integrates Lyapunov optimization with the deep deterministic policy gradient(DDPG)algorithm.The proposed approach transforms the latency minimization problem under long-term cost constraints into a time-slot-level optimization task.By leveraging DDPG to handle dynamic optimization in high-dimensional and nonlinear environments,the algorithm achieves efficient allocation and coordination of system resources.Simulation results demonstrate that,compared with existing approaches,the proposed algorithm is more effective in reducing overall system latency while satisfying cost constraints.关键词
移动边缘计算/深度强化学习/边缘缓存/深度确定性策略梯度算法Key words
mobile edge computing/deep reinforcement learning/edge caching/deep deterministic policy gradient(DDPG)algorithm分类
信息技术与安全科学引用本文复制引用
唐宏,燕星芮,施杰,刘子兴..基于深度强化学习的基站协作与服务缓存研究[J].重庆邮电大学学报（自然科学版）,2026,38(3):391-400,10.基金项目
国家自然科学基金资助项目(61971080) National Natural Science Fondation of China(61971080) （61971080）
