Mechanism analysis and feature extraction algorithm of communication emitter fingerprint

Abstract

A fingerprint extraction method based on the fingerprint mechanism (FM) of a communication emitter is proposed to solve the shortcomings of existing emitter identification technology. First, the Taylor polynomial model of the communication emitter fingerprint is established by analyzing the mechanism of the emitter fingerprint. Second, the polynomial coefficients containing the emitter fingerprints are defined with focus on single-frequency and narrow-band signal. Identifiable features are extracted through parameter estimation. Theoretical analysis shows that the FM algorithm focuses on the fingerprint essence of communication emitter, and the effects of modulations are overcome. Experimental results show that the identification performance of the FM algorithm is better than that of three other algorithms due to the high anti-noise performance, inner-class clustering and inter-class scattering of extracted features.