Method for detecting unknown modulation mode of radar signals based on generative adversarial network

Abstract

The invention relates to a method for detecting an unknown modulation mode of radar signals based on a generative adversarial network, and belongs to the field of electronic recognition. The method comprises the following steps of collecting all radar signals sequentially from a radar signal database; obtaining time-frequency images of radar signals through Cui-Williams distribution and short-timeFourier transform by using time-frequency transform; performing parameters training on the generative adversarial network with the time-frequency images, combined with deep learning, and preserving discriminator parameters; constructing discriminators based on the discriminator parameter, and discriminating the radar signals by combining the respective discriminators; optimizing the weight parameters of each discriminator by using the improved crow search algorithm and combined with the idea of the group intelligent optimization algorithm to obtain the best result of discrimination; finally retaining the parameters of the discriminators and the weight parameters, and detecting the received radar signals. The method provided by the invention combines time-frequency transform, deep learningand the improved crow search algorithm to realize stable, fast and high-precision detection of radar radiation source signals with unknown modulation mode, and has broad application prospects.

Description

technical field [0001] The invention relates to a method for detecting an unknown modulation mode of a radar signal based on a generated confrontation network, which belongs to the field of electronic identification. Background technique [0002] The identification of radar radiation source signal modulation mode is an important link in electronic identification. In identification, more information can be obtained to grasp the initiative. However, with the emergence of new radars, the modulation methods of radar signals are becoming more and more complex and changeable. The types of radar signal modulation methods in the known library are limited, and more radar signal modulation methods are unknown. The existing modulation method identification methods cannot detect radar emitter signals with unknown modulation methods well, so it is urgent and meaningful to find new detection methods. For the detection of unknown signals, there is generally outlier detection, but outlier ...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to provide a radar signal unknown modulation method detection method based on generative adversarial networks in order to solve the problems of poor stability, long time consumption, and low precision in the detection of radar emitter signals with unknown modulation methods in existing detection methods

Images