Automatic identification method for radar signal modulation type

Abstract

According to the automatic identification method for the radar signal modulation type, the identification mode is simple and efficient, and training is stable. The method is realized through the following technical scheme: constructing an end-to-end signal modulation type identification network model by adopting a one-dimensional convolutional neural network, and taking a real part, an imaginary part and a phase of a complex signal as signal representation information; the method comprises the following steps: sampling a complex signal sample by a radar, carrying out normalization processing on the complex signal sample, building a modulation type identification network model formed by a residual-attention convolution block, a full connection layer and a classifier, and outputting a predicted modulation type through feature extraction, feature integration and probability mapping; after a modulation type identification network model is determined, network model parameters such as a loss function, an optimizer, a learning rate and the like are set, the network model is trained and the network parameters are updated by taking minimization of the loss function as a target through a back propagation algorithm, and an offline training-online detection signal modulation mode classifier is obtained, so that radar signal modulation type identification is realized.

Description

technical field [0001] The invention relates to a radar signal modulation type identification method based on a residual-attention convolutional neural network, especially a method capable of accurately identifying the modulation type of a given signal. Background technique [0002] Radar emitter signal identification is through processing and analyzing the intercepted radar signal, and mining information such as radar type, status and function, so as to master its information. The identification of radar signal modulation type is the core of radiation source identification. Not only that, radar radiation source signal identification obtains the unique attributes of the signal by analyzing the pulse characteristics of different modulation waveform signals, and realizes radar modulation type identification. Therefore, fast and efficient radar signal modulation recognition plays a vital role in electromagnetic spectrum domain countermeasures, which directly affects the accurac...

AI Technical Summary

Problems solved by technology

In general, in the case of small samples, problems such as poor convergence and reduced timeliness increase as the network deepens

In practical applications, the application of deep networks is limited by conditions such as sample size and computing resources

In addition, in the training process of deep convolutional neural network, most of the existing radar signal modulation type recognition methods ignore the correlation between the output features of the convolutional layer, and do not take into account the different contributions of different output feature maps to target recognition, and As the number of network layers increases, there will be a problem of gradient disappearance

[0006] At present, the existing one-dimensional convolutional neural network radar signal modulation type recognition methods based on deep learning mostly use the amplitude value of the complex signal as a one-dimensional vector or concatenate the real and imaginary parts of the complex signal into a one-dimensional vector as one-dimensional convolution The single-channel input of the neural network does not take full advantage of the characteristics of the complex signal, namely: real part, imaginary part and phase information

In addition, due to insufficient deep representation ability, the depth of the network is very large, and its deep feature extraction process is very time-consuming, which is not suitable for classification tasks in low SNR environments and multiple signal types

Images