Intelligent Decision Method of Slope Perturbing Based on Q-Learning for Anti-Deception Jamming

In the dynamic deception jamming scenes, it is difficult for radar to successfully suppress jamming and detect targets by randomly changing frequency modulation (FM) slope. In order to make radar learn the jamming strategy of parameters change, suppress jamming in real time and optimize the perturbing strategy of FM slope, an intelligent decision method of FM slope perturbing based on Q-learning is proposed. Firstly, we design a jamming suppression method based on signal processing. In order to ensure that it can successfully distinguish jamming and target after suppression, we design constraints of FM slope optimization to improve the correlation performance of waveforms. Then, we design decision basis, establish decision model based on Q-learning, and deduce the objective function of this problem. Finally, we combine the suppression method and Q-learning to design two intelligent decision algorithms. The establishment of knowledge base and on-line decision of radar anti-jamming are realized. Simulation results show that the proposed method has low decision time, high decision accuracy and strong anti-jamming performance compared with existing methods.