Abstract: For the problem of the poor performance of traditional specific emitter identification methods in low signal to noise ratio (SNR) environments, a method of specific emitter identification for the aerial target is proposed. Empirical mode decomposition and variational mode decomposition are employed to obtain modal components of different frequencies of the signals, and multi-scale entropy of each modal component is taken as features. The principal component analysis is used to reduce the dimensions of the features, and the support vector machine is used as a classifier to identify the specific emitter identification. Simulation results show that the proposed method has better recognition and anti-noise performance for fine features such as phase noise, frequency drift and harmonic distortion than the traditional methods.

Key words: specific emitter identification (SEI), fine features, mode decomposition, multi-scale permutation entropy, support vector machine (SVM)