基于特征融合的跳频信号射频指纹识别技术

射频指纹识别(RFID)是一种物理层身份认证的方法，是电子对抗中一个重要且基本的研究方向，为现代战争提供情报信息等方面发挥着重要作用；为了提升在电子战复杂环境下RFID的准确率，同时解决在跳频信号片段长度有限致使稳态特征难以提取的问题，提出了一种基于信号多个维度特征融合与深度卷积网络提取特征的智能识别技术，改进了传统的星座图特征提取方法并提取了信号的双谱、星座图和希尔伯特-黄变换 (HHT)时频谱进行特征融合，并设置了不同信噪比和不同输入条件下的对照实验来证明该方法的有效性和鲁棒性；相比于传统的识别方法，该方法运算量小，且提升了在各信噪比下识别准确率，在正常室外环境下对六部相移键控(PSK)类跳频电台的识别准确率达到了99.29%。

RFID Technology of FH Signal Based on Feature Fusion

Radio frequency fingerprint identification (RFID) is a method of physical layer identity authentication. It is an important and basic research direction in electronic countermeasures, and plays an important role in providing information for modern warfare; In order to improve the accuracy of RFID in the complex environment of electronic warfare, and solve the problem that it is difficult to extract steady-state features due to the limited length of frequency hopping (FH) signal segments, an intelligent recognition technology based on multi dimension feature fusion of signals and deep convolution network feature extraction is proposed. The traditional constellation feature extraction method is improved, and the bispectrum, constellation and spectrum of Hilbert Huang transform (HHT) of signals are extracted for feature fusion, The effectiveness and robustness of this method are proved by setting up contrast experiments under different signal-to-noise ratio (SNR) and different input conditions; Compared with the traditional recognition method, this method has less computation and improves the recognition accuracy under various signal-to-noise ratios. The recognition accuracy of six PSK type FH radios in normal outdoor environment reaches 99.29%.