Radar target identification based on adaptive joint time-frequency processing in high frequency domain

A new radar target feature extraction method based on adaptive joint time-frequency processing is presented for radar target recognition in the high frequency domain. After applying the adaptive Gaussian representation to the range profile, some of the decomposing parameters are extracted as feature vectors used for identification. Comparing with range profiles, the extracted feature vectors proposed by this method increase the identification information and decrease the computation and storing burden significantly when compared to the adaptive spectrogram. The physical meaning is also more definite. This method is used for identifying six kinds of aircraft scale models. The experiment results show that the new feature extraction method is practical.