Radar signal intra-pulse modulation identification method based on deep learning

Abstract

The invention belongs to the technical field of radar signal processing, and discloses a radar signal intra-pulse modulation identification method based on deep learning. The method comprises the following steps: S1, carrying out filtering and de-noising preprocessing on radar sampling signals; S2, performing Cohen type time-frequency distribution processing on the preprocessed sampling signals to obtain a time-frequency image; S3, after the time-frequency image is processed, inputting the time-frequency image into a trained DCNN-C network model, and automatically judging the type of input radar intra-pulse modulation signals through the network model to complete identification, wherein the DCNN-C network model comprises a DCNN network and a classification network spliced with the DCNN network. According to the method, the Choi-Williams time-frequency distribution processing is carried out by using a double-sphere kernel function, so that the cross term suppression effect on the radar signals is better, and the signal robustness characteristic is more obvious; background denoising processing is carried out on a time-frequency image by using the DCNN network, information loss of signal energy caused by time-frequency preprocessing can be effectively avoided, so that the accuracy of radiation source identification is improved, and the identification method is simple, practical, and effective.

Description

technical field [0001] The invention belongs to the technical field of radar signal processing, and in particular relates to a radar signal intrapulse modulation recognition method based on deep learning. Background technique [0002] Intra-pulse modulation and identification of radar signals play an important role in informationized battles. Conducting electronic countermeasures against the enemy's radar communication system and detecting local intelligence are of great significance to enhancing the ability of military weapons and equipment to resist the enemy. Cognitive radio and public safety and other fields have high application value. [0003] For radar signal intra-pulse modulation identification, the traditional method is to extract time of arrival (Time Of Arrival, TOA), angle of arrival (Angle Of Arrival, AOA), carrier frequency (Carrier Frequency, CF), pulse amplitude (Pulse Amplitude, PA) and pulse width (Pulse Width, PW) sequence parameters to form a pulse desc...

AI Technical Summary

Problems solved by technology

[0005] The present invention provides a radar signal intrapulse modulation recognition method based on deep learning, aiming to solve the technical problems of low signal processing efficiency, loss of original signal information and low signal recognition accuracy in the prior art

Images