基于DCGAN的雷达辐射源信号个体识别算法OACSTPCD
Individual Identification Algorithm for Radar Emitter Signal Based on DCGAN
雷达辐射源个体识别技术通过提取雷达细微特征判定载体身份属性,是电子对抗领域的热点研究方向之一.通过深度学习识别雷达辐射源指纹是当前的主流方法,然而训练网络需要大量的数据样本,当数据样本不足时,容易造成识别准确率不高等突出问题.基于此,提出了一种基于深度卷积对抗网络(Deep Convolutional Generative Adversarial Networks,DCGAN)的雷达辐射源信号个体识别算法,首先采用双谱切片对信号进行特征提取,然后构建了基于DCGAN的识别网络,最后通过实采数据验证了算法的有效性.实验结果表明,在样本缺失较为严重的条件下,所提出的算法能实现小样本条件下的雷达辐射源识别,识别准确率达到90％,完全满足日常需求.
The radar emitter individual recognition technology determines the carrier identity attributes by ex-tracting radar subtle features,which is one of the hot research directions in the field of electronic countermeas-ures.It is a mainstream method to identify the fingerprint of radar emitter by deep learning.However,the train-ing network requires a large number of data samples,and when the data samples are insufficient,the recogni-tion accuracy is not high.Based on this,an individual recognition algorithm of radar emitter signal based on deep convolutional generative adversarial networks(DCGAN)is proposed.Firstly,bispectral slice is used to ex-tract the features of the signal.Then DCGAN-based recognition network is constructed.Finally,the validity of the algorithm is verified by real data.The experimental results show that under the condition of severe sample loss,the proposed algorithm can recognize radar emitter under small sample conditions,with a recognition accu-racy of 90％,fully meeting the daily needs.
王程昱;凌青;闫文君
海军航空大学信息融合研究所,山东烟台 264001
计算机与自动化
雷达个体识别小样本DCGAN深度学习
radar individual recognitionsmall sampleDCGANdeep learning
《测控技术》 2024 (007)
17-22,64 / 7
国家自然科学基金面上项目(62371465);山东省青创团队资助(2022KJ084)
评论