%0 Journal Article
%T 基于DDPG的基站节能决策优化方法
Base Station Energy-Saving Decision Optimization Method Based on DDPG
%A 杨忱逊
%A 高月红
%A 杨鸿文
%J Hans Journal of Wireless Communications
%P 27-38
%@ 2163-3991
%D 2025
%I Hans Publishing
%R 10.12677/hjwc.2025.152004
%X 在通信网络中，当用户通信需求随时间降低时，基站可执行两种节能决策：一是直接关断低负载基站，将用户的服务转移至其他基站；二是通过功率分级调节逐级降低基站的能耗，同时维持通信服务的稳定性。本研究基于深度确定性策略梯度(Deep Deterministic Policy Gradient, DDPG)的基站节能决策优化方法，将综合考虑用户通信服务质量和系统能耗，选择性能更优的节能决策。实验结果表明，相较于基站保持高功率运行的基线实验，文章提出的两种节能决策能够使系统能耗和时延的综合性能分别提升47.5%和56.4%。对比两种节能决策，直接关断表现出更快的收敛特性，而功率分级调节则在能耗和时延的整体性能上更具优势，比关断决策高出16.9%。
In communication networks, when the user communication demand decreases over time, base stations can implement two energy-saving decisions: one involves directly shutting down low-load base stations and transferring user services to other operational base stations; the other entails gradually reducing the energy consumption of base stations through hierarchical power adjustment, ensuring that communication services remain stable. In this paper, the base station energy-saving decision optimization method based on Deep Deterministic Policy Gradient (DDPG) will comprehensively consider the user communication service quality and system energy consumption and select the optimal energy-saving decision. The experimental results show that, compared with the baseline experiment in which base stations maintain high power operation, the two energy-saving decisions proposed in this paper can improve the comprehensive performance of system energy consumption and delay by 47.5% and 56.4%, respectively. When comparing the two energy-saving decisions, direct shutdown shows faster convergence, while hierarchical power adjustment has more advantages in the overall performance of energy consumption and delay, representing a 16.9% improvement over the shutdown decision.
%K 基站节能，
%K 强化学习，
%K 决策优化
Base Station Energy Saving
%K Reinforcement Learning
%K Decision Optimization
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=112755