Joint Optimization of Frequency Selection and Transmit Power for Radar Anti-Jamming using Reinforcement Learning

With the widespread deployment of radio frequency equipment, spectrum coexistence between radar and commu-nication systems has become increasingly common. Thus, it is essential for radar systems to avoid mutual interference between systems in congested spectral environments. Moreover, mainlobe jamming transmitted by self-protection jammers with interception mechanism poses a significant threat due to their rapid response capabilities to radar pulses. To address these challenges, this paper proposes a reinforcement learning based joint optimization of frequency selection and transmit power (RL-JOFSTP) approach for countering communication interference and mainlobe jamming. First, the coexistence environment and the self-protection jammer are introduced. The anti-jamming strategy learning problem is then modeled as a Markov Decision Process (MDP), with proximal policy optimization (PPO) employed to solve the joint optimization problem. The frequency selection strategy enables the radar to operate in unoccupied frequency bands, while the transmit power strategy reduces the interception probability by the self-protection jammer. Finally, simulation results show that the proposed RL-JOFSTP method significantly enhances radar anti-jamming performance and finds the optimal countermeasure strategy.