Anti-Jamming Technique for IRS Aided JRC System in Mobile Vehicular Networks

Undesired jamming launched by malicious jammers can attack authorized communications, which is viewed as one of the critical challenges in vehicular networks. In this paper, in order to handle the problem, anti-jamming communication driven by reinforcement learning is studied in intelligent reflecting surface (IRS)-aided vehicular networks. The system sum transmission rate is optimized by joint designing the transmit beamforming at the roadside unit (RSU) and the reflection coefficients at the IRS. An anti-jamming strategy based on combining annealing bias-priority experience replay method and twin delayed deep deterministic policy gradient (TD3) technique is developed to handle the formulated challenging non-convex problem. The proposed strategy is employed to train the replay buffer in TD3, which can eliminate the deviation under the distribution change and has the advantages of fast convergence and is not easy to fall into local optima. Numerical results confirm that our proposed strategy can enhance the sum rate of multiple vehicular users and ensure radar sensing capability of RSU compared with the existing methods.