杂波和干扰条件下基于强化学习的机载雷达波形设计  

Airborne Radar Waveform Design Based on Reinforcement Learning under Clutter and Jamming Conditions

作　　者：郑泽新 李伟[1] 邹鲲[1] 李艳福 ZHENG Zexin;LI Wei;ZOU Kun;LI Yanfu(Information and Navigation School,Air Force Engineering University,Xi’an 710077,China)

机构地区：[1]空军工程大学信息与导航学院,西安710077

出　　处：《火力与指挥控制》2023年第5期39-45,52,共8页Fire Control & Command Control

基　　金：国家自然科学基金面上项目(62271500);陕西省自然科学基金面上资助项目(2020JM-343)。

摘　　要：针对复杂电磁环境机载雷达智能抗干扰问题,提出一种基于马尔可夫决策过程(markov decision processes,MDP)的机载雷达波形设计方法。为实现最优决策,建立雷达和干扰MDP博弈模型,融合利用目标、杂波、噪声、雷达和干扰信号等多维度电磁信息,设置信号和干扰噪声比为奖励函数;基于贝尔曼方程和策略迭代法计算信号频域最优策略,通过迭代变换法设计时域最优波形,并采用目标检测概率衡量算法性能。仿真结果表明,和线性调频信号、跳频信号相比,该方法设计波形具有更好的环境适应性和抗干扰能力,在此基础上提高了机载雷达目标检测概率。Aiming at the intelligent anti-jamming problem of airborne radar in the complex electromagnetic environment,a waveform design method of airborne radar based on the Markov Decision Process(MDP)is proposed.In order to achieve optimal decision-making,a game model of radar and jamming MDP is established,and multi-dimensional electromagnetic information such as target,clutter,noise,radar and jamming signals,etc.is integrated and used,and the signal-to-interference-plus-noise ratio is set as the reward function;the optimal strategy in the frequency domain of the signal is calculated based on Bellman equation and strategy iteration method,the optimal waveform in the time domain is designed through the iterative transformation method,and the target detection probability is used to measure the performance of the algorithm.The simulation results show that compared with the LFM signal and frequency hopping signal,the waveform designed by this method has better environmental adaptability and anti-jamming ability,and on this basis,the detection probability of airborne radar targets is improved.

关 键 词：雷达波形设计 抗干扰 博弈 马尔可夫决策过程 

分 类 号：TN958[电子电信—信号与信息处理]