Engineering a Lightweight Deep Joint Source-Channel-Coding-Based Semantic Communication System

Deep joint source-channel coding (DeepJSCC) has emerged as a novel technology in semantic communication, coinciding with the increasing demand for the edge devices in the Internet of Things (IoT). Consequently, the deployment of DeepJSCC on edge devices has become a crucial research direction. Howev...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:IEEE internet of things journal Ročník 12; číslo 1; s. 458 - 471
Hlavní autoři: Zhang, Weihan, Wu, Shaohua, Meng, Siqi, He, Jinghang, Zhang, Qinyu
Médium: Journal Article
Jazyk:angličtina
Vydáno: Piscataway IEEE 01.01.2025
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Témata:
ISSN:2327-4662, 2327-4662
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Abstract Deep joint source-channel coding (DeepJSCC) has emerged as a novel technology in semantic communication, coinciding with the increasing demand for the edge devices in the Internet of Things (IoT). Consequently, the deployment of DeepJSCC on edge devices has become a crucial research direction. However, DeepJSCC faces challenges related to channel fading. Moreover, implementing DeepJSCC on the edge devices poses challenges due to the constrained computational resources as well as the compatibility issue between DeepJSCC and digital systems. In this article, we devote to engineering the DeepJSCC system deployed on the edge devices. First, we propose a method named DeepJSCC with Ensemble learning (DeepJSCC-ES) to resist the channel fading. Then, we present a pruning algorithm called the DeepJSCC signal-to-noise ratio (SNR)-adaptive pruning method (DJSAP) to make the DeepJSCC network lightweight, reducing the computational demands on the edge nodes. Further, we propose a method called the simulated fixed-point quantization training based on soft quantization function (SFPQSQ) to tackle the compatibility issue between DeepJSCC and digital systems. Finally, we deploy the whole DeepJSCC system on the edge devices and conduct experiments to test the DeepJSCC system. The results of simulations show that the proposed DeepJSCC-ES system outperforms the baseline DeepJSCC, particularly excelling in low SNR conditions. Furthermore, the parameter size of the pruned model using DJSAP is compressed by 93.37% while the average structural similarity index metric (SSIM) decreases only by 0.92% compared with the baseline DeepJSCC. Additionally, the SFPQSQ works better than the ordinary quantization methods in tackling the compatibility issue between DeepJSCC and digital systems. The experiment results also show that our proposed system can serve as a feasible solution for practical deployment on the edge devices.
AbstractList Deep joint source-channel coding (DeepJSCC) has emerged as a novel technology in semantic communication, coinciding with the increasing demand for the edge devices in the Internet of Things (IoT). Consequently, the deployment of DeepJSCC on edge devices has become a crucial research direction. However, DeepJSCC faces challenges related to channel fading. Moreover, implementing DeepJSCC on the edge devices poses challenges due to the constrained computational resources as well as the compatibility issue between DeepJSCC and digital systems. In this article, we devote to engineering the DeepJSCC system deployed on the edge devices. First, we propose a method named DeepJSCC with Ensemble learning (DeepJSCC-ES) to resist the channel fading. Then, we present a pruning algorithm called the DeepJSCC signal-to-noise ratio (SNR)-adaptive pruning method (DJSAP) to make the DeepJSCC network lightweight, reducing the computational demands on the edge nodes. Further, we propose a method called the simulated fixed-point quantization training based on soft quantization function (SFPQSQ) to tackle the compatibility issue between DeepJSCC and digital systems. Finally, we deploy the whole DeepJSCC system on the edge devices and conduct experiments to test the DeepJSCC system. The results of simulations show that the proposed DeepJSCC-ES system outperforms the baseline DeepJSCC, particularly excelling in low SNR conditions. Furthermore, the parameter size of the pruned model using DJSAP is compressed by 93.37% while the average structural similarity index metric (SSIM) decreases only by 0.92% compared with the baseline DeepJSCC. Additionally, the SFPQSQ works better than the ordinary quantization methods in tackling the compatibility issue between DeepJSCC and digital systems. The experiment results also show that our proposed system can serve as a feasible solution for practical deployment on the edge devices.
Author Zhang, Weihan
Meng, Siqi
Wu, Shaohua
Zhang, Qinyu
He, Jinghang
Author_xml – sequence: 1
  givenname: Weihan
  orcidid: 0009-0000-8630-2234
  surname: Zhang
  fullname: Zhang, Weihan
  email: 200210730@stu.hit.edu.cn
  organization: School of Electronics and Information Engineering, Harbin Institute of Technology (Shenzhen), Shenzhen, China
– sequence: 2
  givenname: Shaohua
  orcidid: 0000-0002-6950-0594
  surname: Wu
  fullname: Wu, Shaohua
  email: hitwush@hit.edu.cn
  organization: Guangdong Provincial Key Laboratory of Aerospace Communication and Networking Technology, Harbin Institute of Technology (Shenzhen), Shenzhen, China
– sequence: 3
  givenname: Siqi
  orcidid: 0000-0002-9273-1163
  surname: Meng
  fullname: Meng, Siqi
  email: mengsiqi@stu.hit.edu.cn
  organization: School of Electronics and Information Engineering, Harbin Institute of Technology (Shenzhen), Shenzhen, China
– sequence: 4
  givenname: Jinghang
  orcidid: 0009-0008-0337-8881
  surname: He
  fullname: He, Jinghang
  email: 190210406@stu.hit.edu.cn
  organization: School of Electronics and Information Engineering, Harbin Institute of Technology (Shenzhen), Shenzhen, China
– sequence: 5
  givenname: Qinyu
  orcidid: 0000-0001-9272-0475
  surname: Zhang
  fullname: Zhang, Qinyu
  email: zqy@hit.edu.cn
  organization: Guangdong Provincial Key Laboratory of Aerospace Communication and Networking Technology, Harbin Institute of Technology (Shenzhen), Shenzhen, China
BookMark eNpNkE1PAjEQhhuDiYj8ABMPTTwv9mPbXY66okJIOIBXm9IOUMK2uF1i-PeWwIFLp5M878zkuUcdHzwg9EjJgFIyfJmMZ4sBIywf8FxyKdgN6jLOiiyXknWu_neoH-OWEJJigg5lF_2M_Np5gMb5NdZ46tab9g9OL34H2ONJcL7F83BoDGTVRnsPu6wKNuHZm45g8Rxq7VtncBXq-uCd0a0LHs-PsYX6Ad2u9C5C_1J76PtjtKi-sunsc1y9TjPDctlm2lLNwRpJiOBLu5JlyagdasFFUS4htYIJu4RcWEkZA84TDgxEaXJqKOM99Hyeu2_C7wFiq7bpZJ9WKk7zQrCCM5ooeqZME2JsYKX2jat1c1SUqJNJdTKpTibVxWTKPJ0zDgCueFlyWUr-D9chcOA
CODEN IITJAU
Cites_doi 10.1109/COMST.2022.3223224
10.1007/s11277-022-10111-7
10.1109/TCCN.2022.3151935
10.1109/WCNC55385.2023.10119015
10.1109/JSAIT.2022.3231042
10.1109/JIOT.2022.3201017
10.1109/TCCN.2019.2919300
10.1109/JSAIT.2020.2987203
10.1109/LWC.2023.3256006
10.1109/GLOBECOM54140.2023.10436878
10.1109/LSP.2022.3184251
10.1109/JIOT.2017.2767608
10.1109/ICC45041.2023.10278961
10.1109/ICCC54389.2021.9674297
10.1109/MWC.010.2300180
10.23919/JCC.2022.03.007
10.1109/tcomm.2024.3386577
10.1109/SPAWC48557.2020.9154306
10.1109/JSAC.2017.2760186
10.1109/comst.2024.3416309
10.1109/GCWkshps56602.2022.10008561
10.1109/ICSIP57908.2023.10270968
10.1109/LCOMM.2023.3329533
10.1109/CCNC51644.2023.10060488
10.1109/TCCN.2023.3326302
10.1109/mnet.2024.3418554
10.1109/JSAC.2020.3018807
10.1109/IOTM.001.2300167
10.1109/LWC.2022.3204837
10.1109/LSP.2021.3113827
10.1109/ICASSP43922.2022.9746335
10.1109/JSAC.2023.3288238
10.1109/ICASSP39728.2021.9414037
10.1109/mwc.014.2300492
10.1109/ICC45041.2023.10279541
10.1109/MCOM.004.2200819
10.1109/ICC45041.2023.10278612
10.1109/PIMRC56721.2023.10293843
10.1109/JIOT.2022.3220677
10.1109/SPAWC53906.2023.10304536
10.1109/JSAC.2022.3221991
ContentType Journal Article
Copyright Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2025
Copyright_xml – notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2025
DBID 97E
RIA
RIE
AAYXX
CITATION
7SC
8FD
JQ2
L7M
L~C
L~D
DOI 10.1109/JIOT.2024.3463652
DatabaseName IEEE All-Society Periodicals Package (ASPP) 2005-present
IEEE All-Society Periodicals Package (ASPP) 1998-Present
IEEE Electronic Library (IEL)
CrossRef
Computer and Information Systems Abstracts
Technology Research Database
ProQuest Computer Science Collection
Advanced Technologies Database with Aerospace
Computer and Information Systems Abstracts – Academic
Computer and Information Systems Abstracts Professional
DatabaseTitle CrossRef
Computer and Information Systems Abstracts
Technology Research Database
Computer and Information Systems Abstracts – Academic
Advanced Technologies Database with Aerospace
ProQuest Computer Science Collection
Computer and Information Systems Abstracts Professional
DatabaseTitleList
Computer and Information Systems Abstracts
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Computer Science
EISSN 2327-4662
EndPage 471
ExternalDocumentID 10_1109_JIOT_2024_3463652
10683686
Genre orig-research
GrantInformation_xml – fundername: Guangdong Basic and Applied Basic Research Foundation
  grantid: 2022B1515120002
  funderid: 10.13039/501100001321
– fundername: National Natural Science Foundation of China
  grantid: 62027802
  funderid: 10.13039/501100001809
– fundername: Major Key Project of PCL
  grantid: PCL2024A01
GroupedDBID 0R~
6IK
97E
AAJGR
AARMG
AASAJ
AAWTH
ABAZT
ABJNI
ABQJQ
ABVLG
AGQYO
AHBIQ
AKJIK
AKQYR
ALMA_UNASSIGNED_HOLDINGS
ATWAV
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
EBS
IFIPE
IPLJI
JAVBF
M43
OCL
PQQKQ
RIA
RIE
4.4
AAYXX
AGSQL
CITATION
EJD
7SC
8FD
JQ2
L7M
L~C
L~D
ID FETCH-LOGICAL-c246t-ad1a3edc60053bdf68821d9a53578be688525dbe45d6122e33a3ee2e58c41c123
IEDL.DBID RIE
ISICitedReferencesCount 0
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001381411000015&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 2327-4662
IngestDate Mon Jun 30 13:00:36 EDT 2025
Sat Nov 29 01:44:10 EST 2025
Wed Aug 27 02:30:24 EDT 2025
IsPeerReviewed false
IsScholarly true
Issue 1
Language English
License https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html
https://doi.org/10.15223/policy-029
https://doi.org/10.15223/policy-037
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c246t-ad1a3edc60053bdf68821d9a53578be688525dbe45d6122e33a3ee2e58c41c123
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ORCID 0009-0000-8630-2234
0009-0008-0337-8881
0000-0001-9272-0475
0000-0002-6950-0594
0000-0002-9273-1163
PQID 3147527321
PQPubID 2040421
PageCount 14
ParticipantIDs crossref_primary_10_1109_JIOT_2024_3463652
proquest_journals_3147527321
ieee_primary_10683686
PublicationCentury 2000
PublicationDate 2025-01-01
PublicationDateYYYYMMDD 2025-01-01
PublicationDate_xml – month: 01
  year: 2025
  text: 2025-01-01
  day: 01
PublicationDecade 2020
PublicationPlace Piscataway
PublicationPlace_xml – name: Piscataway
PublicationTitle IEEE internet of things journal
PublicationTitleAbbrev JIoT
PublicationYear 2025
Publisher IEEE
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Publisher_xml – name: IEEE
– name: The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
References ref13
ref35
ref12
ref34
ref15
ref37
ref14
ref36
ref31
ref30
ref11
ref33
ref10
ref32
ref2
ref1
ref17
ref39
ref16
ref38
ref19
ref18
Wang (ref27) 2022
ref24
ref23
ref26
ref25
ref20
ref42
ref41
ref22
ref21
ref28
Molchanov (ref44) 2016
ref29
ref8
ref7
Liu (ref45) 2021
ref9
ref4
Han (ref43) 2015
ref3
ref6
ref5
ref40
References_xml – ident: ref1
  doi: 10.1109/COMST.2022.3223224
– ident: ref23
  doi: 10.1007/s11277-022-10111-7
– year: 2021
  ident: ref45
  article-title: AdaPruner: Adaptive channel pruning and effective weights inheritance
  publication-title: arXiv:2109.06397
– ident: ref33
  doi: 10.1109/TCCN.2022.3151935
– ident: ref35
  doi: 10.1109/WCNC55385.2023.10119015
– ident: ref38
  doi: 10.1109/JSAIT.2022.3231042
– ident: ref20
  doi: 10.1109/JIOT.2022.3201017
– ident: ref8
  doi: 10.1109/TCCN.2019.2919300
– ident: ref12
  doi: 10.1109/JSAIT.2020.2987203
– ident: ref21
  doi: 10.1109/LWC.2023.3256006
– ident: ref14
  doi: 10.1109/GLOBECOM54140.2023.10436878
– ident: ref40
  doi: 10.1109/LSP.2022.3184251
– ident: ref18
  doi: 10.1109/JIOT.2017.2767608
– ident: ref31
  doi: 10.1109/ICC45041.2023.10278961
– ident: ref29
  doi: 10.1109/ICCC54389.2021.9674297
– ident: ref42
  doi: 10.1109/MWC.010.2300180
– ident: ref7
  doi: 10.23919/JCC.2022.03.007
– year: 2022
  ident: ref27
  article-title: SNN-SC: A spiking semantic communication framework for feature transmission
  publication-title: arXiv:2210.06836
– ident: ref39
  doi: 10.1109/tcomm.2024.3386577
– ident: ref26
  doi: 10.1109/SPAWC48557.2020.9154306
– ident: ref16
  doi: 10.1109/JSAC.2017.2760186
– ident: ref2
  doi: 10.1109/comst.2024.3416309
– ident: ref10
  doi: 10.1109/GCWkshps56602.2022.10008561
– ident: ref32
  doi: 10.1109/ICSIP57908.2023.10270968
– ident: ref36
  doi: 10.1109/LCOMM.2023.3329533
– year: 2015
  ident: ref43
  article-title: Deep compression: Compressing deep neural networks with pruning, trained quantization and Huffman coding
  publication-title: arXiv:1510.00149
– ident: ref37
  doi: 10.1109/CCNC51644.2023.10060488
– ident: ref25
  doi: 10.1109/TCCN.2023.3326302
– ident: ref4
  doi: 10.1109/mnet.2024.3418554
– ident: ref19
  doi: 10.1109/JSAC.2020.3018807
– ident: ref22
  doi: 10.1109/IOTM.001.2300167
– ident: ref30
  doi: 10.1109/LWC.2022.3204837
– ident: ref24
  doi: 10.1109/LSP.2021.3113827
– ident: ref15
  doi: 10.1109/ICASSP43922.2022.9746335
– ident: ref9
  doi: 10.1109/JSAC.2023.3288238
– ident: ref28
  doi: 10.1109/ICASSP39728.2021.9414037
– year: 2016
  ident: ref44
  article-title: Pruning convolutional neural networks for resource efficient transfer learning
  publication-title: arXiv:1611.06440
– ident: ref5
  doi: 10.1109/mwc.014.2300492
– ident: ref6
  doi: 10.1109/ICC45041.2023.10279541
– ident: ref13
  doi: 10.1109/MCOM.004.2200819
– ident: ref41
  doi: 10.1109/ICC45041.2023.10278612
– ident: ref11
  doi: 10.1109/PIMRC56721.2023.10293843
– ident: ref17
  doi: 10.1109/JIOT.2022.3220677
– ident: ref34
  doi: 10.1109/SPAWC53906.2023.10304536
– ident: ref3
  doi: 10.1109/JSAC.2022.3221991
SSID ssj0001105196
Score 2.334865
Snippet Deep joint source-channel coding (DeepJSCC) has emerged as a novel technology in semantic communication, coinciding with the increasing demand for the edge...
SourceID proquest
crossref
ieee
SourceType Aggregation Database
Index Database
Publisher
StartPage 458
SubjectTerms Algorithms
Coding
Communications systems
Compatibility
Deep joint source-channel coding (DeepJSCC)
Devices
Digital systems
edge devices
Ensemble learning
Fading
Image reconstruction
Internet of Things
Internet of Things (IoT)
Lightweight
lightweight model
Mathematical models
quantization
Quantization (signal)
Rician channels
Rician fading channel
Semantics
Signal to noise ratio
Training
Weight reduction
Title Engineering a Lightweight Deep Joint Source-Channel-Coding-Based Semantic Communication System
URI https://ieeexplore.ieee.org/document/10683686
https://www.proquest.com/docview/3147527321
Volume 12
WOSCitedRecordID wos001381411000015&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVIEE
  databaseName: IEEE Electronic Library (IEL)
  customDbUrl:
  eissn: 2327-4662
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0001105196
  issn: 2327-4662
  databaseCode: RIE
  dateStart: 20140101
  isFulltext: true
  titleUrlDefault: https://ieeexplore.ieee.org/
  providerName: IEEE
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3PS8MwFA5uePDi_DFxOiUHT0Lm0qZpc9Tp0DGmsAk7WdLkFQbajq3Tf98k7XBDPHhraQrlfX393nvp-x5CV5AIxQEYSVikCFPAiQykJioEIUWqAz91SA_D0SiaTsVL1azuemEAwP18Bh176Pbyda5WtlRmPJxHPo94DdXCMCybtX4KKtRGI7zauaRdcTN4ep6YDNBjHd_KYgXeFve4YSq_vsCOVvqNfz7QAdqv4kd8WwJ-iHYgO0KN9WwGXLnqMXrbEBrEEg9tDv7lyqD4HmCOB_ksK_DYle6JbTHI4J30cstk5M4wm8Zj-DBWnym81UOCS4nzJnrtP0x6j6SapUCUx3hBpKbSB6249bpEp9xE1lQLGVi1mwTMaeAFOgEWaBPzeOD7Zjl4EESKUWXo7QTVszyDU4RTJpIkpZbIzDWTMUmTJQrGJOWJCehkC12vrRzPS8mM2KUaXRFbSGILSVxB0kJNa9aNhaVFW6i9BiauvGoZ-5SFVjDOo2d_3HaO9jw7oNfVSNqoXixWcIF21WcxWy4u3QvzDfn6wKs
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LS8QwEB58gV58i6ur5uBJiJo0yTZHn6y6roIreLKkyRQWtLvoqn_fJO2iIh68tTSFMl-n38yk8w3ALubaKkRBc5FaKiwqaqRx1LZQG104mRQR6U6r200fHvRt3awee2EQMf58hvvhMO7lu4F9C6Uy7-EqTVSqJmFaCsFZ1a71VVJhIR5R9d4lO9QHlxc3PZ8DcrGfBGEsyX-wTxyn8usbHInlfOGfj7QI83UESY4qyJdgAstlWBhPZyC1s67A4zepQWJIJ2ThH7EQSk4Rh-Ry0C9H5C4W72loMijxiZ4MApfRY89tjtzhs7d735IfXSSkEjlfhfvzs95Jm9bTFKjlQo2occwk6KwKfpe7QvnYmjltZNC7ydGfSi5djkI6H_VwTBK_HDnK1ApmPcGtwVQ5KHEdSCF0nhcsUJm_5nMm4_NELYRhKvchnWnA3tjK2bASzchisnGoswBJFiDJakgasBrM-m1hZdEGNMfAZLVfvWYJE60gGcfZxh-37cBsu3fdyToX3atNmONhXG-smDRhavTyhlswY99H_deX7fjyfAIrnsPy
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Engineering+a+Lightweight+Deep+Joint+Source-Channel-Coding-Based+Semantic+Communication+System&rft.jtitle=IEEE+internet+of+things+journal&rft.au=Zhang%2C+Weihan&rft.au=Wu%2C+Shaohua&rft.au=Meng%2C+Siqi&rft.au=He%2C+Jinghang&rft.date=2025-01-01&rft.pub=IEEE&rft.eissn=2327-4662&rft.volume=12&rft.issue=1&rft.spage=458&rft.epage=471&rft_id=info:doi/10.1109%2FJIOT.2024.3463652&rft.externalDocID=10683686
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=2327-4662&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=2327-4662&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=2327-4662&client=summon