Specific Emitter Identification Algorithm Based on Electromagnetic Gene Feature Extraction

As the complexity of electronic warfare continues to grow, the role of specific emitter identification (SEI) technology in electronic reconnaissance becomes increasingly significant. Traditional methods depend on the surface characteristics of received signals, which have limitations such as instability and a lack of physical correlation. To address these issues, this article introduces an SEI algorithm grounded in electromagnetic gene feature extraction. Initially, emitter gene features are defined using biological gene theory, and a Wiener-enhanced memory polynomial (WEMP) model is proposed. The WEMP model integrates the strengths of the Wiener model and the memory polynomial (MP) model to improve their nonlinear modeling capabilities. Then, the Wigner-Ville distribution (WVD) is employed to extract signal parameters for high-precision signal reconstruction. Furthermore, a differential evolution (DE) algorithm is used to optimize an error-adaptive iterative feature extraction process, which significantly enhances robustness at low signal-to-noise ratios (SNRs). Experimental results compare the proposed method with traditional time-frequency analysis using data from a hardware-in-the-loop (HIL) platform. Results achieved indicate that the proposed method's recognition rate reaches 97.61% at an SNR of 1 dB, which is over 10% higher than that of empirical and variational modal decomposition (VMD) methods.