Radar Radiating Signal Feature Extraction and Recognition
Wu Shaopeng · Jisuanji fangzhen · 2011
Radar emitter classification and identification are composite task,which plays an important role in military automated control and command system.Under the radar system for the complex characteristics of radar signals,in order to improve the radar emitter signal recognition rate of individual,a wavelet packet transform is used for feature extraction of the band energy of the signal which can reflect the unintentional modulation on pulse(UMOP).Then the mixed kernel function SVM with strong generalization ability and learning ability is used for classification and recognition,which is compared with the classification capability of Gaussian Kernel Support Vector Machine.A high recognition rate is obtained.Simulation results prove that this method obtain the desired results.