Radiation source modulation mode identification method based on SEResNet-LSTM

Abstract

The invention discloses a radiation source modulation mode identification method based on SEResNet-LSTM, and the method comprises the steps: carrying out the signal preprocessing of an original communication signal data set, and obtaining the IQ data of a radiation source to be identified; the method comprises the following steps: constructing an SEResNet-LSTM network, wherein the SEResNet-LSTM network comprises a multi-scale feature extraction module, an LSTM fusion network and a softmax layer which are cascaded; constructing a training data set, and training the SEResNet-LSTM network by using the training data set; and inputting the IQ data of the radiation source into the trained SEResNet-LSTM network model to obtain the modulation mode type of the original communication signal. According to the method, the ResNet network and the LSTM network are combined, a channel attention mechanism is introduced, space and time features of signals are considered, automatic feature extraction is performed on original IQ data, deep features of the signals are mined, and the requirement of automatic modulation recognition is met.

Description

technical field [0001] The invention belongs to the technical field of communication signal processing, and in particular relates to a radiation source modulation mode identification method based on SEResNet-LSTM. Background technique [0002] With the continuous innovation of information technology and new equipment, the modulation methods of communication radiation sources are increasingly diversified, and various modulation methods emerge in an endless stream, and the identification of radiation source modulation methods is the basis of communication reconnaissance and radio signal cognition. Therefore, modulation identification technology The importance of electronic reconnaissance and electromagnetic spectrum countermeasures has become increasingly prominent. [0003] At present, common radiation source modulation identification methods at home and abroad include traditional Bayesian decision theory-based, feature extraction-based and deep learning-based modulation iden...

AI Technical Summary

Problems solved by technology

[0005] For example, directly use CNN network and RNN network (Recurrent Neural Networks, recurrent neural network) to automatically extract effective signal baseband IQ (in-phase branch signal and quadrature branch signal) component features, and based on high-order cumulants (High-order Cumulants , HOC) SVM (Support Vector Machine, SVM) modulation recognition method achieves a better recognition effect, but this method only uses a simple neural network, and the signal feature extraction ability is limited

Images