Anti-interference Intelligent Decision-making Technology Based on Deep Reinforcement Learning

In recent years, with the rapid development of wireless communication technology in the military field and the proliferation of the number of various types of wireless communication devices, the battlefield electromagnetic environment has become more and more complex, which puts forward higher requirements for the anti-interference ability of communication systems. In order to enhance the reliability and adaptability of communication systems in complex electromagnetic environments, this paper combines Deep Q-Learning (DQN) with Long Short-Term Memory (LSTM), and proposes a communication anti-jamming intelligent decision- making algorithm based on DQN-LSTM, which realizes adaptive frequency use. In this method, dynamic s-greedy strategy and LSTM are utilized to improve the convergence and stability of DQN. Simulation results show better anti-jamming performance compared to traditional anti-jamming methods and traditional DQN decision-making algorithms.