Q-Learning Based Adaptive Frequency Hopping Strategy Under Probabilistic Jamming

It is difficult to ensure stability and communication efficiency of traditional anti-jamming frequency hopping communication system in the frequency domain probabilistic jamming environment. Therefore, this paper proposes an adaptive frequency hopping communication scheme under probabilistic jamming. Based on the probabilistic jamming model, the practical frequency hopping (FH) strategy and average payoff is obtained using Q-Learning algorithm and is compared with the theoretical optimal average payoff obtained by the greedy algorithm. Simulation results show that our proposed algorithm is fast to converge, has fewer frequency hops compared with existing anti-jamming FH strategy, and the average payoff is close to the theoretical optimal frequency hopping strategy.