Emitter identification based on improved Variational modal decomposition at low SNR

Abstract

Individual identification of radiation source is a technique based on extracting RF signal features. A new RF fingerprint extraction method is proposed for low SNR environment and low individual identification rate of radiation sources. The VMD algorithm is used to decompose the signal, and the autocorrelation function is used to optimize it to strengthen the suppression of the modal aliasing phenomenon after the signal decomposition. The decomposed modes were analyzed by the Teager energy operator, and the energy and box dimension were extracted as fingerprint features, which were input into the support vector machine (SVM) for classification and recognition. The simulation results show the effectiveness of using VMD to extract fingerprint features under low SNR environment. And the proposed emitter identification scheme can improve the recognition rate of emitter in the environment with low signal-tonoise ratio (SNR) -5~0dB.