3D convolutional neural networks based automatic modulation classification in the presence of channel noise
Abstract Automatic modulation classification is a task that is essentially required in many intelligent communication systems such as fibre‐optic, next‐generation 5G or 6G systems, cognitive radio as well as multimedia internet‐of‐things networks etc. Deep learning (DL) is a representation learning...
Main Authors: | , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Wiley
2022-03-01
|
Series: | IET Communications |
Online Access: | https://doi.org/10.1049/cmu2.12269 |
_version_ | 1811337051337916416 |
---|---|
author | Rahim Khan Qiang Yang Inam Ullah Ateeq Ur Rehman Ahsan Bin Tufail Alam Noor Abdul Rehman Korhan Cengiz |
author_facet | Rahim Khan Qiang Yang Inam Ullah Ateeq Ur Rehman Ahsan Bin Tufail Alam Noor Abdul Rehman Korhan Cengiz |
author_sort | Rahim Khan |
collection | DOAJ |
description | Abstract Automatic modulation classification is a task that is essentially required in many intelligent communication systems such as fibre‐optic, next‐generation 5G or 6G systems, cognitive radio as well as multimedia internet‐of‐things networks etc. Deep learning (DL) is a representation learning method that takes raw data and finds representations for different tasks such as classification and detection. DL techniques like Convolutional Neural Networks (CNNs) have a strong potential to process and analyse large chunks of data. In this work, we considered the problem of multiclass (eight classes) classification of modulated signals, which are, Binary Phase Shift Keying, Quadrature Phase Shift Keying, 16 and 64 Quadrature Amplitude Modulation corrupted by Additive White Gaussian Noise, Rician and Rayleigh fading channels using 3D‐CNN architectures in both frequency and spatial domains while deploying three approaches for data augmentation, which are, random zoomed in/out, random shift and random weak Gaussian blurring augmentation techniques with a cross‐validation (CV) based hyperparameter selection statistical approach. Simulation results testify the performance of 10‐fold CV without augmentation in the spatial domain to be the best while the worst performing method happens to be 10‐fold CV without augmentation in the frequency domain and we found learning in the spatial domain to be better than learning in the frequency domain. |
first_indexed | 2024-04-13T17:49:21Z |
format | Article |
id | doaj.art-030f9a512f9d49e7adc2e0fcbbe53765 |
institution | Directory Open Access Journal |
issn | 1751-8628 1751-8636 |
language | English |
last_indexed | 2024-04-13T17:49:21Z |
publishDate | 2022-03-01 |
publisher | Wiley |
record_format | Article |
series | IET Communications |
spelling | doaj.art-030f9a512f9d49e7adc2e0fcbbe537652022-12-22T02:36:47ZengWileyIET Communications1751-86281751-86362022-03-0116549750910.1049/cmu2.122693D convolutional neural networks based automatic modulation classification in the presence of channel noiseRahim Khan0Qiang Yang1Inam Ullah2Ateeq Ur Rehman3Ahsan Bin Tufail4Alam Noor5Abdul Rehman6Korhan Cengiz7School of Electronics and Information Engineering Harbin Institute of Technology Harbin ChinaSchool of Electronics and Information Engineering Harbin Institute of Technology Harbin ChinaCollege of Internet of Things (IoT) Engineering Hohai University (HHU) Changzhou Campus Changzhou ChinaDepartment of Electrical Engineering Government College University Lahore PakistanSchool of Electronics and Information Engineering Harbin Institute of Technology Harbin ChinaCISTER Research Centre ISEP Politécnico do Porto PortugalDepartment of Computer Science and Engineering Kyungpook National University Daegu South KoreaDepartment of Electrical ‐ Electronics Engineering Trakya University Edirne TurkeyAbstract Automatic modulation classification is a task that is essentially required in many intelligent communication systems such as fibre‐optic, next‐generation 5G or 6G systems, cognitive radio as well as multimedia internet‐of‐things networks etc. Deep learning (DL) is a representation learning method that takes raw data and finds representations for different tasks such as classification and detection. DL techniques like Convolutional Neural Networks (CNNs) have a strong potential to process and analyse large chunks of data. In this work, we considered the problem of multiclass (eight classes) classification of modulated signals, which are, Binary Phase Shift Keying, Quadrature Phase Shift Keying, 16 and 64 Quadrature Amplitude Modulation corrupted by Additive White Gaussian Noise, Rician and Rayleigh fading channels using 3D‐CNN architectures in both frequency and spatial domains while deploying three approaches for data augmentation, which are, random zoomed in/out, random shift and random weak Gaussian blurring augmentation techniques with a cross‐validation (CV) based hyperparameter selection statistical approach. Simulation results testify the performance of 10‐fold CV without augmentation in the spatial domain to be the best while the worst performing method happens to be 10‐fold CV without augmentation in the frequency domain and we found learning in the spatial domain to be better than learning in the frequency domain.https://doi.org/10.1049/cmu2.12269 |
spellingShingle | Rahim Khan Qiang Yang Inam Ullah Ateeq Ur Rehman Ahsan Bin Tufail Alam Noor Abdul Rehman Korhan Cengiz 3D convolutional neural networks based automatic modulation classification in the presence of channel noise IET Communications |
title | 3D convolutional neural networks based automatic modulation classification in the presence of channel noise |
title_full | 3D convolutional neural networks based automatic modulation classification in the presence of channel noise |
title_fullStr | 3D convolutional neural networks based automatic modulation classification in the presence of channel noise |
title_full_unstemmed | 3D convolutional neural networks based automatic modulation classification in the presence of channel noise |
title_short | 3D convolutional neural networks based automatic modulation classification in the presence of channel noise |
title_sort | 3d convolutional neural networks based automatic modulation classification in the presence of channel noise |
url | https://doi.org/10.1049/cmu2.12269 |
work_keys_str_mv | AT rahimkhan 3dconvolutionalneuralnetworksbasedautomaticmodulationclassificationinthepresenceofchannelnoise AT qiangyang 3dconvolutionalneuralnetworksbasedautomaticmodulationclassificationinthepresenceofchannelnoise AT inamullah 3dconvolutionalneuralnetworksbasedautomaticmodulationclassificationinthepresenceofchannelnoise AT ateequrrehman 3dconvolutionalneuralnetworksbasedautomaticmodulationclassificationinthepresenceofchannelnoise AT ahsanbintufail 3dconvolutionalneuralnetworksbasedautomaticmodulationclassificationinthepresenceofchannelnoise AT alamnoor 3dconvolutionalneuralnetworksbasedautomaticmodulationclassificationinthepresenceofchannelnoise AT abdulrehman 3dconvolutionalneuralnetworksbasedautomaticmodulationclassificationinthepresenceofchannelnoise AT korhancengiz 3dconvolutionalneuralnetworksbasedautomaticmodulationclassificationinthepresenceofchannelnoise |