A Specific Emitter Identification Method Based on Time-Frequency Feature Extraction
Wenlong Dong, Yuqi Wang, Guang‐Cai Sun, Mengdao Xing · 2023
With the rapid growth of the Internet of Things (IoT), fundamental security measures of wireless networks have become a basic requirement. Aiming at the identification of wireless transmitters with the same parameters, this paper proposes a specific emitter identification (SEI) method based on time-frequency feature extraction. Received signals go through preprocessing, i.e., multipath effect estimation and Doppler frequency compensation, to mitigate the channel effect. Then the time-frequency spectrum is generated and a time-frequency feature extraction network is constructed to achieve feature extraction and identification task. Real-world data are used to verify the effectiveness of the proposed method. The overall identification accuracy for stationary emitters reaches 92.9%. Besides, the proposed preprocessing method improves moving emitter identification accuracy by 12%.