Radar emitter fingerprint recognition based on bispectrum and SURF feature
Naixin Kang, Minghao He, Jun Han, Bing-qie Wang · 2016
Radar emitter identification via a collection of received radar signals is a subject of great importance in modern electronic warfare. Aimed at radar emitter fingerprint recognition, bispectrum theory and SURF feature are applied in the paper and have achieved good performance. Due to the fingerprint characteristics reflected from bispectrum transform such as phase noise, fine distinction of signals from diverse sources is obvious in bispectrum projection. Thus bispectrum transform is capable of classifying signals of the same modulation type and similar parameters from different radar emitters. In order to make best use of the distinction, we convert bispectrum projection to grayscale, from which to extract SURF feature and recognize them by feature matching. Simulation results have proved the validity of the method when signals are even with unconspicious difference.