A Dynamic Game Strategy for Radar Screening Pulsewidth Allocation Against Jamming Using Reinforcement Learning

Radio frequency (RF) screening is a practical antijamming technique. Determining screening pulse width is key to the performance of RF screening. In this article, we investigate the problem of radar screening pulsewidth allocation against a cognitive jammer, which also seeks a strategy to best jam the radar. As both the radar and the jammer are taken as intelligent agents learning to optimize their respective performance, dynamic game theory is employed and the problem is constructed as an extensive-form game. A strategy learning approach based on reinforcement learning is proposed to approximate the Nash equilibrium (NE). Simulation results verify that the proposed learning approach outperforms existing ones and that the learned strategies approximate the NE.