摘要: |
无人机对于无线信道的依赖性和无线传播环境的开放性,导致其通信易受到恶意的电磁干扰。针对其中恶意的信道跟随干扰,在感知干扰信道信息的基础上,将无人机的发射功率和信道选择策略建模为马尔科夫决策过程(Markov Decision Process,MDP),利用强化学习算法对该通信系统的抗干扰方法进行智能优化,提出了基于赢或快学习策略爬山算法(Win or Learn Fast Policy Hill-climbing,WoLF-PHC)的抗干扰算法。仿真结果证明,所提算法能够将用户干信比降低至0.1以下,将用户可达速率在初始值基础上提升14%,与Q学习算法和PHC算法相比具有更好的抗干扰传输性能。 |
关键词: 无人机(UAV) 抗干扰 电磁干扰感知 强化学习 |
DOI:10.20079/j.issn.1001-893x.230401001 |
|
基金项目:国家自然科学基金资助项目(62071485,62271503,62001513);江苏省自然科学基金项目(BK20201334,BK20201334,BK20200579,BK20231485) |
|
An Electromagnetic Jamming Sensing and Anti-jamming Transmission Method of UAV Based on Reinforcement Learning |
LI Boyang,LIU Yang,WAN Nuotian,XU Kui,XIA Xiaochen,ZHANG Yueyue,ZHANG Mi |
((College of Communications Engineering,Army Engineering University of PLA,Nanjing 210007,China)) |
Abstract: |
The dependence on wireless channel and the openness of wireless channel make unmanned aerial vehicle(UAV) vulnerable to malicious electromagnetic jammings.To combat channel-following jamming from jammers,a reinforcement learning-based anti-jamming strategy is proposed based on the perception of jamming spectrum information.The power control and channel access strategy of UAV is modeled as Markov Decision Process(MDP),and the anti-jamming strategy of the communication system is intelligently optimized by Reinforcement Learning Algorithm.An anti-jamming algorithm based on Win or Learn Fast Policy Hill-climbing(WoLF-PHC) algorithm is proposed.The simulation results prove that the proposed algorithm can reduce the user’s Interference-to-Signal Ratio(ISR) less than 0.1,and increase the user’s achievable rate by 14% on the basis of the initial value.Compared with Q-learning Algorithm and PHC Algorithm,it has better anti-interference transmission performance. |
Key words: unmanned aerial vehicle(UAV) anti-jamming electromagnetic jamming sensing reinforcement learning |