A hidden anti-jamming method based on deep reinforcement learning

Most of the current anti-jamming algorithms for wireless communications only consider how to avoid jamming attacks, but ignore that communication waveform or frequency action may be obtained by jammers. Although existing methods can guarantee temporary effects, long-term performance these depressed when intelligent jammers are capable learning from historical activities. Aiming at this issue, a hidden method based on idea reducing jammer's sense probability is proposed. Firstly, sensing jammer calculating correlation between actions and user. Later, deep reinforcement framework designed, which aims not maximizing throughput also minimizing action's Finally, algorithm proposed, links instantaneous return with quality users jammer. The simulation result shows proposed avoids being sensed improves its compared considers avoidance.