基于分层Q学习的联合抗干扰算法

Joint Anti-jamming Algorithm Based on Hierarchical Q Learning

针对智能干扰威胁下的跨层抗干扰通信问题,提出一种基于分层Q学习的联合抗干扰学习算法。根据用户与干扰机之间的路由信道选择问题构建分层Stackelberg博弈模型,干扰机选择最佳干扰信道实施干扰,用户与干扰机进行路由信道博弈,选择最佳路由及信道实现通信。仿真结果表明,与固定路由-随机信道选择算法、随机路由-最佳信道选择算法和随机路由-随机信道选择算法相比,该算法具有更好的抗干扰性能。

Aiming at the cross-layer anti-jamming communication problem under the threat of intelligent interference,a joint anti-jamming learning algorithm based on hierarchical Q learning is proposed.The problem of routing selection and channel allocation between users and intelligent jammer is modeled as a hierarchical Stackelberg game.In the routing-chanel game between users and jammers,the intelligent jammer chooses the best channel for jamming,while users select the best route and channels for communication.Simulation results show that compared with the Fixed-routing-random-channel Selection Algorithm(FRSA),Random-routing-optimal-channel Selection Algorithm(ROSA) and Random-routing-random-channel Selection Algorithm(RRSA),the proposed algorithm has better anti-jamming capacity.