A novel collaborative jamming method based on decoy strategy

As a tool for real-time online interaction with the environment, reinforcement learning is widely used in communication confrontation. We proposed a novel cooperative jamming method based on decoy strategy to fight the anti-jamming user based on Q-Learning algorithm. With the same power consumption, the proposed strategy is greatly improved compared with traditional jamming methods such as frequency scanning jamming and fixed-band jamming. Simulation results show that the strategy prevents the intelligent users from learning the complete jamming policy, i. e., its strategy cannot converge. With the jamming mode switched every 400 steps, compared with traditional jamming, the average jamming accuracy increases 36.9%, the user average throughput reducing 53.7%.