Radio Signal Feature Extraction Method Based on Attention Deep Network

A radio signal and deep network technology, applied in the field of radio signal feature extraction based on attention deep network, can solve the problems of radio signal identification, radio signal feature extraction, and complex feature extraction methods, etc., to achieve the characteristics of radio signals Simple extraction method, simplified steps, overcoming the effect of feature extraction

Active Publication Date: 2019-12-24
XIDIAN UNIV
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The disadvantage of this method is that although this method proposes a method for identifying the modulation mode of cognitive radio signals under the condition of low signal-to-noise ratio, this method requires frequency domain conversion and complex mathematical operations when extracting signal features , requires a lot of prior knowledge, highly relies on manual feature extraction, and the feature extraction method is complex
The disadvantage of this method is that although this method proposes a radio feature extraction method based on neural network to realize signal classification, it cannot perform effective feature extraction on radio signals, so that it cannot realize the identification of radio signal modulation and coding methods , and this method requires wavelet analysis of one-dimensional signals to use neural networks for feature extraction and classification recognition

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
  • Radio Signal Feature Extraction Method Based on Attention Deep Network
  • Radio Signal Feature Extraction Method Based on Attention Deep Network
  • Radio Signal Feature Extraction Method Based on Attention Deep Network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0038] The invention will be further described below in conjunction with the drawings.

[0039] Reference attached figure 1 , To further describe the specific steps of the present invention.

[0040] Step 1. Construct a coded and modulated joint radio signal.

[0041] Each received radio signal information sequence is subjected to four channel coding in sequence to generate different coded signals.

[0042] The four types of channel coding refer to Hamming code channel coding, 216 non-systematic convolutional code channel coding with one-half rate, 216 non-systematic convolutional code channel coding with two-thirds rate, and quarter Three-rate 432 non-systematic convolutional code channel coding.

[0043] Each coded signal is sequentially subjected to six modulations to obtain a coded-modulated joint radio signal.

[0044] The six types of modulation refer to binary phase shift keying modulation, quaternary phase shift keying modulation, octal phase shift keying modulation, binary digi...

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 discloses a radio signal feature extraction method based on attention depth network. The implementation steps are: generating a training sample set and a test sample set on the basis of constructing a coded and modulated joint radio signal; constructing an attention mechanism layer for extracting The attention depth network of radio signal features; the training sample set is input into the attention depth network for training, and the trained attention depth network is obtained; the recognition accuracy is obtained by using the test sample set and the trained attention depth network. The invention has the advantages of strong universality, no need for manual feature extraction, redundant information removal, low complexity, accurate and stable classification results, and can be used for cognitive recognition and other processing of subsequent radio signals.

Description

Technical field [0001] The present invention belongs to the field of communication technology, and further relates to a radio signal feature extraction method based on the attention depth network in the field of signal processing technology. The present invention simulates the visual selective attention mechanism in the information processing mechanism of the human brain, and can automatically extract radio signal characteristics and quickly screen out high-value characteristics in a complex electromagnetic environment for subsequent cognitive recognition of radio signals. Compared with the existing deep learning model, the present invention not only has the characteristics of low calculation complexity, small parameter scale, and easy hardware implementation, but also can obtain accurate radio signal modulation and coding mode identification under a lower signal-to-noise ratio. Background technique [0002] Radio signal feature extraction is the basic step of radio signal proces...

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
Patent Type & Authority Patents(China)
IPC IPC(8): H04L27/00G06N3/08G06N3/04
CPCH04L27/0012G06N3/08G06N3/045
Inventor 杨淑媛王敏吴亚聪焦李成黄震宇王喆李兆达张博闻宋雨萱李治王翰林王俊骁
Owner XIDIAN 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