Wireless signal noise reduction method based on generative adversarial network

Abstract

The invention discloses a wireless signal noise reduction method based on a generative adversarial network. The method comprises the following steps: constructing a wireless signal noise reduction network GAN-G and a wireless signal noise reduction discrimination network GAN-D, where the GAN-G comprises an LSTM, a convolution layer and a time attention module, the input of the time attention module is an original wireless signal with noise added, and the output of the time attention module is a noise-reduced wireless signal; the GAN-D comprising an LSTM and a full connection layer, the input is an original wireless signal and a noise-reduced wireless signal, and the output is a judgment result; training the GAN-G and the GAN-D by using a adversarial training network, to obtain a wireless signal noise reduction model and a wireless signal noise reduction discrimination model; and extracting the wireless signal noise reduction model to process the to-be-noise-reduced wireless signal to obtain a noise-reduced wireless signal. The method can effectively improve the accuracy of wireless signal modulation type identification in a low signal-to-noise ratio interval by carrying out noise reduction processing on the wireless signal data.

Description

technical field [0001] The invention belongs to the application of a deep learning method in the field of wireless signal recognition, and in particular relates to a wireless signal noise reduction method based on a generative confrontation network. Background technique [0002] With its powerful feature learning ability, deep learning technology is widely used in various fields, mainly including computer vision, natural language processing, complex network analysis and wireless signal analysis. Deep learning is a method based on representational learning of data in machine learning. It extracts features from data through a huge neural network, thereby simulating the mechanism of the human brain to interpret data. Typical deep neural networks include convolutional neural networks and recurrent neural networks. Among them, convolutional neural networks are widely used in image classification tasks and target detection tasks due to their powerful feature extraction performance...

AI Technical Summary

Problems solved by technology

Although the wireless signal modulation technology is relatively mature, when the wireless signal contains a lot of noise, the identification of the wireless signal modulation type will become difficult, and the existing identification methods will be affected to varying degrees, so that the accuracy of identification cannot be guaranteed.

Images