Linear Jamming Bandits: Learning to Jam OFDM-Modulated Signals

This work investigates the use of linear reinforce-ment learning to effectively jam an OFDM-modulated victim signal. Prior work has shown improved convergence with the use of linear bandits, a variant of reinforcement learning, to jam single-carrier digital phase-amplitude modulation schemes using time-domain (TD) jamming schemes. However, communication systems today typically employ orthogonal frequency division multiplexing (OFDM) to transmit data, particularly in 4G/5G networks. This work explores the use of linear Thompson Sampling (TS) to efficiently jam OFDM-modulated signals where the jammer may select from single-carrier and OFDM jamming schemes. We show that linear TS performs better than traditional reinforcement learning (UCB-1 algorithm) in terms of maximizing the victim symbol error rate (SER). We also draw novel insights by observing the action states to which the reinforcement learning algorithm converges to.