UniQGAN: Unified Generative Adversarial Networks for Augmented Modulation Classification

Insup Lee; Wonjun Lee

doi:10.1109/LCOMM.2021.3131476

UniQGAN: Unified Generative Adversarial Networks for Augmented Modulation Classification

Insup Lee, Wonjun Lee

Research output: Contribution to journal › Article › peer-review

20 Citations (Scopus)

Abstract

Deep learning has been widely applied to automatic modulation classification (AMC), and there have been many studies on data augmentation techniques using deep generative models to improve performance. However, existing solutions need to train different models independently for each SNR, which leads to undeniable overhead. This letter presents UniQGAN, Unified Generative Adversarial Networks for IQ constellations of various SNRs, requiring a single model training. The proposed method introduces multi-conditions embedding and multi-domains classification to leverage both conditions, i.e., modulation type and SNR. Experimental results show that UniQGAN effectively improves the AMC performance, while the training time is reduced.

Original language	English
Pages (from-to)	355-358
Number of pages	4
Journal	IEEE Communications Letters
Volume	26
Issue number	2
DOIs	https://doi.org/10.1109/LCOMM.2021.3131476
Publication status	Published - 2022 Feb 1

Bibliographical note

Funding Information:
This work was supported by the National Research Foundation (NRF) of Korea grant funded by the Korea Government (MSIT) under Grant 2019R1A2C2088812.

Publisher Copyright:
© 1997-2012 IEEE.

Keywords

Automatic modulation classification
IQ~constellations
generative adversarial networks
single model training

ASJC Scopus subject areas

Modelling and Simulation
Computer Science Applications
Electrical and Electronic Engineering

Access to Document

10.1109/LCOMM.2021.3131476

Cite this

@article{d0a79cdf4b56414c972a25c0b40df086,

title = "UniQGAN: Unified Generative Adversarial Networks for Augmented Modulation Classification",

abstract = "Deep learning has been widely applied to automatic modulation classification (AMC), and there have been many studies on data augmentation techniques using deep generative models to improve performance. However, existing solutions need to train different models independently for each SNR, which leads to undeniable overhead. This letter presents UniQGAN, Unified Generative Adversarial Networks for IQ constellations of various SNRs, requiring a single model training. The proposed method introduces multi-conditions embedding and multi-domains classification to leverage both conditions, i.e., modulation type and SNR. Experimental results show that UniQGAN effectively improves the AMC performance, while the training time is reduced.",

keywords = "Automatic modulation classification, IQ~constellations, generative adversarial networks, single model training",

author = "Insup Lee and Wonjun Lee",

note = "Funding Information: This work was supported by the National Research Foundation (NRF) of Korea grant funded by the Korea Government (MSIT) under Grant 2019R1A2C2088812. Publisher Copyright: {\textcopyright} 1997-2012 IEEE.",

year = "2022",

month = feb,

day = "1",

doi = "10.1109/LCOMM.2021.3131476",

language = "English",

volume = "26",

pages = "355--358",

journal = "IEEE Communications Letters",

issn = "1089-7798",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "2",

}

TY - JOUR

T1 - UniQGAN

T2 - Unified Generative Adversarial Networks for Augmented Modulation Classification

AU - Lee, Insup

AU - Lee, Wonjun

N1 - Funding Information: This work was supported by the National Research Foundation (NRF) of Korea grant funded by the Korea Government (MSIT) under Grant 2019R1A2C2088812. Publisher Copyright: © 1997-2012 IEEE.

PY - 2022/2/1

Y1 - 2022/2/1

N2 - Deep learning has been widely applied to automatic modulation classification (AMC), and there have been many studies on data augmentation techniques using deep generative models to improve performance. However, existing solutions need to train different models independently for each SNR, which leads to undeniable overhead. This letter presents UniQGAN, Unified Generative Adversarial Networks for IQ constellations of various SNRs, requiring a single model training. The proposed method introduces multi-conditions embedding and multi-domains classification to leverage both conditions, i.e., modulation type and SNR. Experimental results show that UniQGAN effectively improves the AMC performance, while the training time is reduced.

AB - Deep learning has been widely applied to automatic modulation classification (AMC), and there have been many studies on data augmentation techniques using deep generative models to improve performance. However, existing solutions need to train different models independently for each SNR, which leads to undeniable overhead. This letter presents UniQGAN, Unified Generative Adversarial Networks for IQ constellations of various SNRs, requiring a single model training. The proposed method introduces multi-conditions embedding and multi-domains classification to leverage both conditions, i.e., modulation type and SNR. Experimental results show that UniQGAN effectively improves the AMC performance, while the training time is reduced.

KW - Automatic modulation classification

KW - IQ~constellations

KW - generative adversarial networks

KW - single model training

UR - http://www.scopus.com/inward/record.url?scp=85120555506&partnerID=8YFLogxK

U2 - 10.1109/LCOMM.2021.3131476

DO - 10.1109/LCOMM.2021.3131476

M3 - Article

AN - SCOPUS:85120555506

SN - 1089-7798

VL - 26

SP - 355

EP - 358

JO - IEEE Communications Letters

JF - IEEE Communications Letters

IS - 2

ER -

UniQGAN: Unified Generative Adversarial Networks for Augmented Modulation Classification

Abstract

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this