Intelligent Reflecting Surface Assisted Anti-Jamming Communications: A Fast Reinforcement Learning Approach

Malicious jamming launched by smart jammers can attack legitimate transmissions, which has been regarded as one of the critical security challenges in wireless communications. With this focus, paper considers use an intelligent reflecting surface (IRS) to enhance anti-jamming communication performance and mitigate interference adjusting elements at IRS. Aiming against a jammer, optimization problem for jointly optimizing power allocation base station (BS) beamforming IRS is formulated while considering quality service (QoS) requirements users. As model behavior are dynamic unknown, fuzzy win or learn fast-policy hill-climbing (WoLF–CPHC) learning approach proposed optimize strategy, where WoLF–CPHC capable quickly achieving optimal policy without knowledge model, state aggregation represent uncertain environment states aggregate states. Simulation results demonstrate that learning-based efficiently improve both IRS-assisted system rate transmission protection level compared with existing solutions.