摘要
现代电磁战环境充斥着大量的人为精心设计的有源干扰,使得依赖技术人员经验的干扰识别技术效果不佳。针对该问题,文中研究了一种基于联合特征平面的残差卷积神经网络(CNN-ResNet)的雷达干扰识别算法,实现雷达干扰类型自动识别分类。该算法通过对雷达回波信号的距离多普勒平面和角度多普勒平面进行预处理,构建联合特征平面,实现对噪声调频干扰、单频干扰、脉冲干扰、转发式干扰和切片干扰的识别和分类。通过仿真表明,该算法可有效提高雷达干扰识别的正确率。
The modern electromagnetic warfare environment is full of plenty of well-designed active jammings.It’s extremely inconvenient to identify these kinds of jammings only by the experience of technicians.To solve this problem,this paper studies a radar jamming identification algorithm based on residual convolution neural network of joint feature planes to realize the automatic identification and classification of the types of radar jamming.This algorithm constructs a joint feature plane and brings about the identification and classification of five different types of radar jamming by pretreating the Doppler planes of distance and angle of radar echo signal.Simulations shows that this algorithm can effectively improve the accuracy of identification of radar jamming.
作者
唐陈
王峰
TANG Chen;WANG Feng(Hohai University,Nanjing 211100,China)
出处
《中国电子科学研究院学报》
北大核心
2022年第1期63-70,共8页
Journal of China Academy of Electronics and Information Technology
关键词
雷达干扰
分类识别
联合特征
卷积神经网络
radar jamming
classification and recognition
joint feature
convolution neural network