Adaptive Jamming Decision Method for Navigation Signals Based on Deep Reinforcement Learning
YUAN Jingmei1， ZHAO Liang2， SUN Zhuoran3，4， XU Zhizhao3， NIU Yalei3
1. Nanjing University of Science and Technology ZiJin College，Nanjing 210023， Jiangsu， China；
2. Shanghai Electro-Mechanical Engineering Institute，Shanghai 201109， China；
3. Beijing Racobit Electronic Information Technology Co.Ltd.，Beijing 100081， China；
4. Radar Technology Research Institute， School of Information and Electronics，
Beijing Institute of Technology， Beijing 100081， China
Abstract：In complex electronic countermeasure environments, efficient selection of jamming signal parameters has remained a critical challenge. This paper proposes an adaptive decision-making method for navigation signal jamming based on deep reinforcement learning, modeling the optimization of jamming signal parameters as a Markov Decision Process (MDP). By rationally designing the state space, action space and reward function, and introducing deep reinforcement learning algorithms, the agent can achieve dynamic adaptive optimization of jamming parameters and autonomously adjust jamming strategies under environmental changes, thereby effectively balancing jamming effectiveness and resource utilization efficiency. Full-link simulation results based on the missile seeker digital simulation platform demonstrate that the proposed method exhibits good convergence and adaptability in scenarios with action spaces of 50 and 100, achieving a final jamming success rate of 99% and frequency band matching success rate of 98%, approaching the performance upper bound of exhaustive search. Further reward function ablation experiments indicate that the designed reward function can effectively guide the agent to achieve reasonable trade-offs among jamming effectiveness, frequency band selection and power consumption, thus forming an efficient, stable and engineering-feasible jamming decision strategy. This research provides new ideas for the development of seeker navigation jamming technology, and can be used to evaluate the anti-jamming capability boundaries of missile navigation signals.
