Carrier Adjustment Algorithm for Mobile Communication Network Based on Deep Reinforcement Learning

To solve the problem of how to accurately and timely adjust the carrier of mobile communication network sectors, a sector expansion (reduction) algorithm based on deep reinforcement learning is proposed. Using Model-based reinforcement learning methods, a multi-model combination of the capacity index probability dynamic model is established. The model is trained using historical data from the real environment, and a virtual environment is constructed based on this. Then, an intelligent agent is built using neural networks, which interact with the virtual environment. The short-unfolding technique is used to generate virtual samples. Finally, the DQN algorithm is used to optimize the strategy of the intelligent agent, enabling it to provide suggestions for sector expansion (reduction) operations. The experimental results show that the carrier adjustment suggestions provided by the trained, intelligent agent achieve a high accuracy rate.