Unlock instant, AI-driven research and patent intelligence for your innovation.

Convolutional neural network modulation identification method based on pseudo constellation diagram

A convolutional neural network and modulation recognition technology, applied in modulation type recognition, neural learning methods, biological neural network models, etc., can solve problems such as ignoring time-related information, and achieve guaranteed accuracy, low complexity, and reduced complexity Effect

Active Publication Date: 2021-12-21
BEIJING UNIV OF POSTS & TELECOMM
View PDF6 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, Convolutional Neural Networks (CNN) methods based on constellation diagrams often ignore the temporal correlation information between adjacent samples.

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
  • Convolutional neural network modulation identification method based on pseudo constellation diagram
  • Convolutional neural network modulation identification method based on pseudo constellation diagram
  • Convolutional neural network modulation identification method based on pseudo constellation diagram

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0030] In this embodiment, a pseudo-constellation diagram is proposed by synthesizing the received time-domain samples and the constellation diagram, in which the former carries the information of time correlation, and the latter carries the information of constellation point distribution, and proposes a method based on convolutional neural network Network (Convolutional Neural Networks, CNN) method. For time correlation, the phase of the sampled value or the in-phase and quadrature components (IQ components) of the signal are used for characterization. For the distribution of constellation points, in order to eliminate the redundancy in the constellation diagram, the pixel values ​​of the grayscale constellation diagram are used for representation. Considering the different degrees of correlation between ranks and columns in the pseudo-constellation, this embodiment further uses an appropriate convolution kernel to optimize the network. Simulation results verify the effectiv...

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 provides a convolutional neural network modulation identification method based on a pseudo constellation diagram. The method comprises the following steps: synthesizing pixel values of a time domain sample and a graying constellation diagram to form the pseudo constellation diagram; inputting the pseudo constellation diagram into a subsequent convolutional neural network; using a preset convolution kernel to optimize the convolutional neural network to form a convolutional neural network model; and performing modulation identification on the to-be-identified signal by using the convolutional neural network model. According to the technical scheme provided by the invention, under the condition that the convolutional neural network model is also used, the problem that a traditional constellation diagram method is weak in MFSK signal recognition capability is solved, so that the overall accuracy of modulation recognition is improved. According to the technical scheme provided by the invention, under the condition that the signal-to-noise ratio is not lower than 10dB, the average recognition accuracy of various PSK, FSK and QAM modulation modes is higher than 0.99, which is far higher than the recognition result of a convolutional neural network based on a traditional constellation diagram. Under the condition of the same input size, the technical scheme provided by the invention has lower complexity.

Description

technical field [0001] The invention relates to the technical field of radio signals, in particular to a convolutional neural network modulation recognition method based on a pseudo-constellation diagram. Background technique [0002] Modulation recognition in the complex electromagnetic field is the basis for the development and wide application of cognitive radio. In recent years, the application of deep learning (DL) to modulation recognition has made great progress. However, Convolutional Neural Networks (CNN) methods based on constellation diagrams often ignore the temporal correlation information between adjacent samples. Contents of the invention [0003] In order to solve the limitations and defects of the prior art, the present invention provides a convolutional neural network modulation recognition method based on a pseudo-constellation diagram, including: [0004] Obtain information on time domain samples and constellation diagrams; [0005] Combining the tim...

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 Applications(China)
IPC IPC(8): G06K9/00G06K9/62G06N3/04G06N3/08H04L27/00
CPCG06N3/08H04L27/0012G06N3/045G06F2218/12G06F18/214
Inventor 徐文波宋世杰王照英王思野
Owner BEIJING UNIV OF POSTS & TELECOMM