Modulated signal time-frequency diagram classification system based on generative adversarial network and operation method thereof

A technology for modulating signals and classification systems, applied in biological neural network models, image analysis, neural learning methods, etc., can solve the problems of insufficient training signal classification neural network data sets, inability to dig deep into signal features, etc., to improve accuracy and the effect of robustness

Active Publication Date: 2020-05-19
SHANDONG UNIV
View PDF3 Cites 18 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] In view of the current ubiquitous problem of insufficient training signal classification neural network data sets and the inability to dig deep into signal features, the present invention provides a modulation signal time-frequency diagram classification system and its ope

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Modulated signal time-frequency diagram classification system based on generative adversarial network and operation method thereof
  • Modulated signal time-frequency diagram classification system based on generative adversarial network and operation method thereof
  • Modulated signal time-frequency diagram classification system based on generative adversarial network and operation method thereof

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0071] A modulated signal time-frequency graph classification system based on GAN. It is a specific application of GAN model in modulation signal classification. It is composed of GAN model and auxiliary classifier model, including IQ signal time-frequency graph Transformation module, generator module, discriminator module;

[0072] The IQ signal time-frequency diagram conversion module and the generator module are all connected to the discriminator module; the discriminator module includes an auxiliary classifier module;

[0073] The IQ signal time-frequency diagram transformation module is used to convert the IQ two-way original modulation signal into a time-frequency diagram form through short-time Fourier transform; the IQ signal time-frequency diagram transformation module runs before generating the confrontational network architecture to generate the confrontational network architecture Including the generator module and the discriminator module; transforming the IQ two-...

Embodiment 2

[0079] According to a kind of modulation signal time-frequency diagram classification system based on generating confrontation network described in embodiment 1, its difference is:

[0080] The IQ signal time-frequency map transformation module includes four parameters for adjustment, including: short-time Fourier transform window type win_cls, short-time Fourier transform window length win_sz, short-time Fourier transform point number nfft, and overlapping point number noverlap. The short-time Fourier transform window type win_cls determines the effect of a short-time Fourier transform, and the corresponding window type can be selected according to different needs. The short-time Fourier transform window type win_cls is a Hanning window, and the window length is and the number of short-time Fourier transform points determines the accuracy of the time-frequency domain. The short-time Fourier transform window length win_sz is 40 signal points, the number of short-time Fourier tr...

Embodiment 3

[0088] The operating method of a modulation signal time-frequency graph classification system based on generating an adversarial network described in Embodiment 2, such as image 3 shown, including the following steps:

[0089] (1) Data preprocessing

[0090] The data set preprocessing is the first step of the entire modulation signal time-frequency diagram classification system. The IQ two-way original modulation signal passes through the IQ signal time-frequency diagram conversion module to convert the IQ two-way original modulation signal into the form of a time-frequency diagram; In order to better explore the deep-level features of the signal, in the IQ signal time-frequency map transformation module, set the short-time Fourier transform window length win_sz to 40 signal points, and set the step to 38 signal points, each time Step by 2 signal points. Such a setting method can better mine the deep-level features of the signal, and provide a better classification basis fo...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention relates to a modulated signal time-frequency diagram classification system based on a generative adversarial network and an operation method of the modulated signal time-frequency diagram classification system. The modulated signal time-frequency diagram classification system comprises an IQ signal time-frequency diagram conversion module, a generator module, a discriminator module and an auxiliary classifier module, wherein the IQ signal time-frequency diagram conversion module converts an original signal into a signal short-time Fourier time-frequency diagram form; the generator module maps an original noise vector and category information input by means of the original noise vector into a corresponding short-time Fourier time-frequency graph; the discriminator module is used for receiving real picture data and picture data generated by a discriminator, and outputting confidence probabilities which are different in input and correspond to each other and are discriminated to be true; and the auxiliary classifier module receives a high-dimensional feature map extracted by a convolution layer and outputs corresponding category information. The operation method is usedfor solving the problem that a deep neural network model for signal classification is trained but the number of data set samples is insufficient.

Description

technical field [0001] The invention relates to a modulation signal time-frequency graph classification system based on a generative confrontation network and an operating method thereof, belonging to the technical field of big data and artificial intelligence. Background technique [0002] In recent years, with the rapid improvement of the performance of hardware-level computing units, deep learning has also demonstrated incredible capabilities in various fields, especially in the fields of image and natural language processing. With the proposal of various new network structures, more and more scholars try to apply the neural network model to the field of signal analysis to classify different types of modulation signals. [0003] As we all know, the training of neural networks requires a large number of data set samples as support, that is, a large number of signal samples are used to train and verify the model. However, in many scenarios, signal acquisition requires a lo...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G06K9/00G06K9/62G06N3/04G06N3/08G06T5/10
CPCG06T5/10G06N3/084G06T2207/20056G06N3/045G06F2218/02G06F2218/08G06F2218/12G06F18/241Y02D30/70
Inventor 王洪君杨晓飞郑庆河王娜许莹胡燕南张德良
Owner SHANDONG UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap