基于深度学习的认知电子对抗技术

Cognitive Electronic Countermeasure Based on Deep Learning

基于现代战场上面临的电磁环境复杂、成功突防难度大的问题,提出一种基于深度学习技术的导弹智能电子对抗系统设计方法。通过将战场上复杂电磁信号的信号类型、威胁等级、干扰策略、效果因子进行量化表征,生成海量数据样本;构建深度学习神经网络模型,并利用生成的数据样本完成模型训练,使模型具备实时感知已知、未知战场射频信号威胁,并生成优化干扰策略的能力;提出一种基于该模型的攻防对抗仿真方法,用于认知电子对抗技术的仿真验证。

Based on the problems of the complex electromagnetic environment and the difficulty of successful penetration for missiles on modern battlefields,a design method of missile intelligent electronic countermeasure system based on deep learning technology is proposed.By quantifying the signal types,threat levels,interference strategies,and effect factors of complex electromagnetic signals on the battlefield,massive data samples are generated;building a deep learning neural network model,and using the generated data samples to complete model training,so that the model has real-time perception known and unknown battlefield RF signal threats and the ability to generate optimized jamming strategies;an attack and defense countermeasure simulation method based on this model is proposed for the simulation verification of cognitive electronic countermeasure technology.