A Lightweight CNN Architecture for Automatic Modulation Classification

Automatic modulation classification (AMC) algorithms based on deep learning (DL) have been widely studied in the past decade, showing significant performance advantage compared to traditional ones. However, the existing DL methods generally behave worse in computational complexity. For this, this pa...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Zhongyong Wang, Dongzhe Sun, Kexian Gong, Wei Wang, Peng Sun
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	automatic modulation classification convolutional neural network depthwise separable convolution feature reconstruction global depthwise convolution Electronics TK7800-8360
Acceso en línea:	https://doaj.org/article/76da91b7b5534820ac74dcded0ea909c
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:76da91b7b5534820ac74dcded0ea909c
record_format	dspace
spelling	oai:doaj.org-article:76da91b7b5534820ac74dcded0ea909c2021-11-11T15:40:13ZA Lightweight CNN Architecture for Automatic Modulation Classification10.3390/electronics102126792079-9292https://doaj.org/article/76da91b7b5534820ac74dcded0ea909c2021-11-01T00:00:00Zhttps://www.mdpi.com/2079-9292/10/21/2679https://doaj.org/toc/2079-9292Automatic modulation classification (AMC) algorithms based on deep learning (DL) have been widely studied in the past decade, showing significant performance advantage compared to traditional ones. However, the existing DL methods generally behave worse in computational complexity. For this, this paper proposes a lightweight convolutional neural network (CNN) for AMC task, where we design a depthwise separable convolution (DSC) residual architecture for feature extraction to prevent the vanishing gradient problem and lighten the computational burden. Besides that, in order to further reduce model complexity, global depthwise convolution (GDWConv) is adopted for feature reconstruction after the last (non-global) convolutional layer. Compared to recent works, the experimental results show that the proposed network can save approximately 70~98% model parameters and 30~99% inference time on two well-known benchmarks.Zhongyong WangDongzhe SunKexian GongWei WangPeng SunMDPI AGarticleautomatic modulation classificationconvolutional neural networkdepthwise separable convolutionfeature reconstructionglobal depthwise convolutionElectronicsTK7800-8360ENElectronics, Vol 10, Iss 2679, p 2679 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	automatic modulation classification convolutional neural network depthwise separable convolution feature reconstruction global depthwise convolution Electronics TK7800-8360
spellingShingle	automatic modulation classification convolutional neural network depthwise separable convolution feature reconstruction global depthwise convolution Electronics TK7800-8360 Zhongyong Wang Dongzhe Sun Kexian Gong Wei Wang Peng Sun A Lightweight CNN Architecture for Automatic Modulation Classification
description	Automatic modulation classification (AMC) algorithms based on deep learning (DL) have been widely studied in the past decade, showing significant performance advantage compared to traditional ones. However, the existing DL methods generally behave worse in computational complexity. For this, this paper proposes a lightweight convolutional neural network (CNN) for AMC task, where we design a depthwise separable convolution (DSC) residual architecture for feature extraction to prevent the vanishing gradient problem and lighten the computational burden. Besides that, in order to further reduce model complexity, global depthwise convolution (GDWConv) is adopted for feature reconstruction after the last (non-global) convolutional layer. Compared to recent works, the experimental results show that the proposed network can save approximately 70~98% model parameters and 30~99% inference time on two well-known benchmarks.
format	article
author	Zhongyong Wang Dongzhe Sun Kexian Gong Wei Wang Peng Sun
author_facet	Zhongyong Wang Dongzhe Sun Kexian Gong Wei Wang Peng Sun
author_sort	Zhongyong Wang
title	A Lightweight CNN Architecture for Automatic Modulation Classification
title_short	A Lightweight CNN Architecture for Automatic Modulation Classification
title_full	A Lightweight CNN Architecture for Automatic Modulation Classification
title_fullStr	A Lightweight CNN Architecture for Automatic Modulation Classification
title_full_unstemmed	A Lightweight CNN Architecture for Automatic Modulation Classification
title_sort	lightweight cnn architecture for automatic modulation classification
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/76da91b7b5534820ac74dcded0ea909c
work_keys_str_mv	AT zhongyongwang alightweightcnnarchitectureforautomaticmodulationclassification AT dongzhesun alightweightcnnarchitectureforautomaticmodulationclassification AT kexiangong alightweightcnnarchitectureforautomaticmodulationclassification AT weiwang alightweightcnnarchitectureforautomaticmodulationclassification AT pengsun alightweightcnnarchitectureforautomaticmodulationclassification AT zhongyongwang lightweightcnnarchitectureforautomaticmodulationclassification AT dongzhesun lightweightcnnarchitectureforautomaticmodulationclassification AT kexiangong lightweightcnnarchitectureforautomaticmodulationclassification AT weiwang lightweightcnnarchitectureforautomaticmodulationclassification AT pengsun lightweightcnnarchitectureforautomaticmodulationclassification
_version_	1718434540097110016

A Lightweight CNN Architecture for Automatic Modulation Classification

Ejemplares similares