Specific Emitter Identification Based on Auto-Correlation and Periodic Feature Extraction

To tackle the challenges associated with identifying long and complex sequence signals, this paper proposes a Specific Emitter Identification (SEI) method based on the Autoformer. In particular, the encoder module is employed to extract periodic correlation features from the I/Q signals, allowing for improved identification accuracy even under complex electromagnetic conditions. By leveraging the Auto-Correlation mechanism, the method effectively captures periodic patterns and enhances the ability to distinguish between different emitters. The approach is evaluated using an ADS-B dataset comprising signals from multiple emitters. Experimental results demonstrate that the proposed method outperforms existing baseline approaches, showcasing its effectiveness in identifying specific signals.