摘要
为解决如何准确、及时地对移动通信网络扇区进行载波调整的问题,提出了一种基于深度强化学习的扇区扩(减)容算法。采用Model-based强化学习方法,建立了容量指标概率动态模型的多模型组合,利用真实环境的历史数据对模型进行训练,并在此基础上构建了虚拟环境。然后用神经网络构建智能体,并使之与虚拟环境互动,采用短展开技术,产生虚拟样本。最后利用虚拟样本,采用DQN算法对智能体进行策略优化,使其给出扇区扩(减)容操作的建议。实验结果表明,训练后的智能体给出的载波调整建议,达到了较高的正确率。
To address the issue of accurately and timely adjusting carriers in mobile communication networks, a sectorexpansion (reduction) algorithm based on deep reinforcement learning (DRL) is proposed. The model-basedreinforcement learning method is utilized to establish a multi-model combination of the probability dynamicmodel of capacity indicators. This model is trained using historical data from the real environment, and a virtualenvironment is constructed based on it. Then, a neural network is used to build an agent, which interacts with thevirtual environment that generates virtual samples using the short rollout technique. Finally, the Deep Q-Network(DQN) algorithm is used to optimize the agent’s strategy using virtual samples, providing suggestions for sectorexpansion (reduction) operations. Experimental results indicate that the trained agent’s carrier adjustmentrecommendations achieve a high level of accuracy.
作者
吕晓阳
沈一飞
吴兵
LV Xiaoyang;SHEN Yifei;WU Bing(Shenzhen Branch of China Mobile Communications Group Guangdong Co.,Ltd.,Shenzhen 518033,China)
出处
《移动通信》
2024年第4期129-134,共6页
Mobile Communications
关键词
移动通信网络
载波调整
深度强化学习
多模型组合
mobile communication network
carrier adjustment
deep reinforcement learning
multi-model combination