摘要
为提高雷达辐射源识别智能水平,提出一种新的基于转换脉冲神经网络进行雷达辐射源调制模式识别的方法。将仿真产生的雷达信号转换为2维时频图,将传统的卷积神经网络(convolutional neural networks, CNN)转化为脉冲神经网络(spiking neuron network, SNN),使用SNN进行雷达辐射源识别。仿真实验结果表明:该方法具有优良的检测精度,当信噪比高于-9 dB时,识别概率可达96%以上。
In order to improve the intelligence level of radar emitter recognition, a new method of radar emitter modulation pattern recognition based on converted spiking neural network is proposed. The simulated radar signal is transformed into a 2D time-frequency map, and the traditional CNN(convolutional neural networks) is transformed into a SNN(spiking neuron network), which is used for radar emitter recognition. The simulation results show that the proposed method has excellent detection accuracy, and the recognition probability can reach more than 96% when the SNR is higher than-9 dB.
作者
李伟
朱卫纲
朱霸坤
杨莹
Li Wei;Zhu Weigang;Zhu Bakun;Yang Ying(Graduate School,Space Engineering University,Beijing 101416,China;Department of Electronic and Optical Engineering,Space Engineering University,Beijing 101416,China)
出处
《兵工自动化》
2022年第7期8-11,20,共5页
Ordnance Industry Automation
关键词
脉冲神经网络
雷达辐射源识别
卷积神经网络
时频转换
spiking neural network
radar emitter identification
convolutional neural networks
time-frequency conversion