Fine Feature Extraction Based on Reverse Analysis of Spectral Regrowth

 In order to identify same-model transmitters in complex wireless environment, this paper researches the fine-feature-extraction method making reverse analysis of spectral regrowth and makes a modification. First, we analyzed spectral regrowth of input signal going through power amplifier. Then we extracted nonlinear parameter as the feature of each power amplifier by power fitting of PSD in selected frequency bands. A modified extraction method was presented in this part. At last, the support vector machine is used in training and classifying. Simulation shows that the method performs well identifying same-model transmitters in Gaussian channel as well as good HF Watterson channel.