Residual Cross-Attention Transformer-Based Multi-User CSI Feedback With Deep Joint Source-Channel Coding

This letter proposes a deep-learning (DL)-based multi-user channel state information (CSI) feedback framework for massive multiple-input multiple-output systems, where the deep joint source-channel coding (DJSCC) is utilized to improve the CSI reconstruction accuracy. Specifically, we design a multi...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE wireless communications letters Jg. 14; H. 8; S. 2481 - 2485
Hauptverfasser: Zhang, Hengwei, Wu, Minghui, Qiao, Li, Liu, Ling, Han, Ziqi, Gao, Zhen
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Piscataway IEEE 01.08.2025
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Schlagworte:
Coding
Correlation
Decoding
deep joint source-channel coding (DJSCC)
Delays
Downlink
Encoding
Feature extraction
Feedback
massive multiple-input multiple-output (MIMO)
multi-user CSI feedback
Noise levels
Residual cross-attention
Training
Transformers
Uplink
Vectors
ISSN:2162-2337, 2162-2345
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Abstract This letter proposes a deep-learning (DL)-based multi-user channel state information (CSI) feedback framework for massive multiple-input multiple-output systems, where the deep joint source-channel coding (DJSCC) is utilized to improve the CSI reconstruction accuracy. Specifically, we design a multi-user joint CSI feedback framework, whereby the CSI correlation of nearby users is utilized to reduce the feedback overhead. Under the framework, we propose a new residual cross-attention transformer architecture, which is deployed at the base station to further improve the CSI feedback performance. Moreover, to tackle the "cliff-effect" of conventional bit-level CSI feedback approaches, we integrated DJSCC into the multi-user CSI feedback, together with utilizing a two-stage training scheme to adapt to varying uplink noise levels. Experimental results demonstrate the superiority of our methods in CSI feedback performance, with low network complexity and better scalability.
AbstractList This letter proposes a deep-learning (DL)-based multi-user channel state information (CSI) feedback framework for massive multiple-input multiple-output systems, where the deep joint source-channel coding (DJSCC) is utilized to improve the CSI reconstruction accuracy. Specifically, we design a multi-user joint CSI feedback framework, whereby the CSI correlation of nearby users is utilized to reduce the feedback overhead. Under the framework, we propose a new residual cross-attention transformer architecture, which is deployed at the base station to further improve the CSI feedback performance. Moreover, to tackle the "cliff-effect" of conventional bit-level CSI feedback approaches, we integrated DJSCC into the multi-user CSI feedback, together with utilizing a two-stage training scheme to adapt to varying uplink noise levels. Experimental results demonstrate the superiority of our methods in CSI feedback performance, with low network complexity and better scalability.
Author Wu, Minghui
Han, Ziqi
Zhang, Hengwei
Qiao, Li
Liu, Ling
Gao, Zhen
Author_xml – sequence: 1
  givenname: Hengwei
  orcidid: 0009-0002-3313-2264
  surname: Zhang
  fullname: Zhang, Hengwei
  email: zhanghengwei2001@163.com
  organization: School of Information and Electronics, Beijing Institute of Technology, Beijing, China
– sequence: 2
  givenname: Minghui
  orcidid: 0000-0001-9195-9556
  surname: Wu
  fullname: Wu, Minghui
  organization: School of Information and Electronics, Beijing Institute of Technology, Beijing, China
– sequence: 3
  givenname: Li
  orcidid: 0000-0003-0586-189X
  surname: Qiao
  fullname: Qiao, Li
  organization: School of Information and Electronics, Beijing Institute of Technology, Beijing, China
– sequence: 4
  givenname: Ling
  orcidid: 0009-0009-0954-6986
  surname: Liu
  fullname: Liu, Ling
  email: liuling@xidian.edu.cn
  organization: Guangzhou Institute of Technology, Xidian University, Guangzhou, China
– sequence: 5
  givenname: Ziqi
  surname: Han
  fullname: Han, Ziqi
  organization: School of Information and Electronics, Beijing Institute of Technology, Beijing, China
– sequence: 6
  givenname: Zhen
  orcidid: 0000-0002-2709-0216
  surname: Gao
  fullname: Gao, Zhen
  email: gaozhen16@bit.edu.cn
  organization: School of Information and Electronics, Beijing Institute of Technology, Beijing, China
BookMark eNpFkMtPAjEQxhuDiYjcPXho4nmxr30dcRXFYEwEwnFT2qkUoYvt7sH_3hKMzmUmme-bx-8S9VzjAKFrSkaUkvJutqpGjLB0xNNcEErPUJ_RjCWMi7T3V_P8Ag1D2JIYGaGMFn20eYdgdSd3uPJNCMm4bcG1tnF44aULpvF78Mm9DKDxa7drbbIM4HE1n-IJgF5L9YlXtt3gB4ADfmmsa_G86byCpNpI5yAObrR1H1fo3MhdgOFvHqDl5HFRPSezt6dpNZ4liom8TYq15KaEQlNBNdOsTNM1B6MLk2bCAJdaKUVVvubcaFoKYtJSGAWEFTSNbT5At6e5B998dRDaehvPcXFlzZnIuChZkUUVOanU8WsPpj54u5f-u6akPiKtI9L6iLT-RRotNyeLBYB_OSU0I3nGfwDVSnPt
CODEN IWCLAF
Cites_doi 10.1109/TCOMM.2022.3217777
10.1109/TAP.2014.2310220
10.48550/ARXIV.1706.03762
10.1109/TCOMM.2024.3524028
10.1109/TCCN.2019.2919300
10.1109/TWC.2020.3043502
10.1109/TPAMI.2023.3275156
10.1109/MWC.008.2200157
10.1109/JSAC.2021.3087269
10.1109/JSAC.2022.3221963
10.1109/MWC.2017.1700147
10.1109/tccn.2024.3486168
10.1109/LWC.2022.3149416
10.1109/TWC.2024.3380638
10.1109/TSP.2014.2324991
10.1109/OJVT.2023.3300464
10.1109/LWC.2018.2818160
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
7SP
8FD
L7M
DOI 10.1109/LWC.2025.3574011
DatabaseName IEEE Xplore (IEEE)
IEEE All-Society Periodicals Package (ASPP) 1998–Present
IEEE Electronic Library (IEL)
CrossRef
Electronics & Communications Abstracts
Technology Research Database
Advanced Technologies Database with Aerospace
DatabaseTitle CrossRef
Technology Research Database
Advanced Technologies Database with Aerospace
Electronics & Communications Abstracts
DatabaseTitleList
Technology Research Database
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Engineering
EISSN 2162-2345
EndPage 2485
ExternalDocumentID 10_1109_LWC_2025_3574011
11016076
Genre orig-research
GrantInformation_xml – fundername: Beijing Nova Program
  funderid: 10.13039/501100005090
– fundername: Natural Science Foundation of Shandong Province; Shandong Province Natural Science Foundation
  grantid: ZR2022YQ62
  funderid: 10.13039/501100007129
– fundername: Natural Science Foundation of China (NSFC)
  grantid: 62471036
  funderid: 10.13039/501100001809
– fundername: Beijing Natural Science Foundation
  grantid: L242011
GroupedDBID 0R~
4.4
5VS
6IK
97E
AAJGR
AARMG
AASAJ
AAWTH
ABAZT
ABQJQ
ABVLG
AGQYO
AGSQL
AHBIQ
AKJIK
AKQYR
ALMA_UNASSIGNED_HOLDINGS
ATWAV
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
EBS
EJD
IES
IFIPE
IPLJI
JAVBF
M43
OCL
RIA
RIE
RNS
AAYXX
CITATION
7SP
8FD
L7M
ID FETCH-LOGICAL-c247t-8ba3f9e8d141d2d2955b3efd8f564fe3adccc1c7b33fd1940f594fce02815e3a3
IEDL.DBID RIE
ISICitedReferencesCount 0
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001547508200005&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 2162-2337
IngestDate Mon Nov 10 11:10:54 EST 2025
Sat Nov 29 07:39:55 EST 2025
Wed Aug 20 06:20:52 EDT 2025
IsPeerReviewed true
IsScholarly true
Issue 8
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-c247t-8ba3f9e8d141d2d2955b3efd8f564fe3adccc1c7b33fd1940f594fce02815e3a3
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ORCID 0000-0003-0586-189X
0009-0009-0954-6986
0000-0002-2709-0216
0000-0001-9195-9556
0009-0002-3313-2264
PQID 3246349286
PQPubID 2040496
PageCount 5
ParticipantIDs proquest_journals_3246349286
crossref_primary_10_1109_LWC_2025_3574011
ieee_primary_11016076
PublicationCentury 2000
PublicationDate 2025-08-01
PublicationDateYYYYMMDD 2025-08-01
PublicationDate_xml – month: 08
  year: 2025
  text: 2025-08-01
  day: 01
PublicationDecade 2020
PublicationPlace Piscataway
PublicationPlace_xml – name: Piscataway
PublicationTitle IEEE wireless communications letters
PublicationTitleAbbrev LWC
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
ref12
ref15
ref14
ref11
ref10
ref2
ref1
ref17
ref18
ref8
ref7
ref9
ref4
ref3
ref6
ref5
(ref16) 2017
References_xml – ident: ref1
  doi: 10.1109/TCOMM.2022.3217777
– volume-title: Study on channel model for frequencies from 0.5 to 100 GHz
  year: 2017
  ident: ref16
– ident: ref17
  doi: 10.1109/TAP.2014.2310220
– ident: ref13
  doi: 10.48550/ARXIV.1706.03762
– ident: ref9
  doi: 10.1109/TCOMM.2024.3524028
– ident: ref14
  doi: 10.1109/TCCN.2019.2919300
– ident: ref10
  doi: 10.1109/TWC.2020.3043502
– ident: ref15
  doi: 10.1109/TPAMI.2023.3275156
– ident: ref4
  doi: 10.1109/MWC.008.2200157
– ident: ref7
  doi: 10.1109/JSAC.2021.3087269
– ident: ref11
  doi: 10.1109/JSAC.2022.3221963
– ident: ref3
  doi: 10.1109/MWC.2017.1700147
– ident: ref6
  doi: 10.1109/tccn.2024.3486168
– ident: ref5
  doi: 10.1109/LWC.2022.3149416
– ident: ref8
  doi: 10.1109/TWC.2024.3380638
– ident: ref2
  doi: 10.1109/TSP.2014.2324991
– ident: ref18
  doi: 10.1109/OJVT.2023.3300464
– ident: ref12
  doi: 10.1109/LWC.2018.2818160
SSID ssj0000601218
Score 2.3744686
Snippet This letter proposes a deep-learning (DL)-based multi-user channel state information (CSI) feedback framework for massive multiple-input multiple-output...
SourceID proquest
crossref
ieee
SourceType Aggregation Database
Index Database
Publisher
StartPage 2481
SubjectTerms Coding
Correlation
Decoding
deep joint source-channel coding (DJSCC)
Delays
Downlink
Encoding
Feature extraction
Feedback
massive multiple-input multiple-output (MIMO)
multi-user CSI feedback
Noise levels
Residual cross-attention
Training
Transformers
Uplink
Vectors
Title Residual Cross-Attention Transformer-Based Multi-User CSI Feedback With Deep Joint Source-Channel Coding
URI https://ieeexplore.ieee.org/document/11016076
https://www.proquest.com/docview/3246349286
Volume 14
WOSCitedRecordID wos001547508200005&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: 2162-2345
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0000601218
  issn: 2162-2337
  databaseCode: RIE
  dateStart: 20120101
  isFulltext: true
  titleUrlDefault: https://ieeexplore.ieee.org/
  providerName: IEEE
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV07T8MwELYAMcDAG1EoyAMLg0tsJ449lkIFCCEEFNiixD6rFVJatYHfj-2kPIQY2CIldqw7J9_d-b47hI6jnCmgmpKcFZEvqm2JNNIS6_DBudhKFIEh93ST3t7Klxd115DVAxcGAELyGXT8ZTjLN2P95kNlp7SuhyYW0WKaipqs9RlQ8YVFWIjnMSoYca9K58eSkTq9ee45Z5AlHZ74HnT0BwyFviq_fsYBYfrr_1zbBlprTEncrXW_iRag3EKr3woMbqPhPcwC2wr3_DJIt6rq9Eb8ODdYYUrOHJIZHKi4ZOC2JO49XOG-g7Ui16_4eVQN8TnABF-PR2WFH0K4n3haQglu4rFHvx006F889i5J01uBaBanFZFFzq0CaWhMDTNMJUnBwTo1JSK2wHOjtaY6LTi3hqo4somKrQZnjtDE3ea7aKkcl7CHsGBaCGmMSjxpTsmCQaS4FEpq7lNAWuhkLupsUpfQyILrEanMqSXzaskatbTQjhft13ONVFuoPVdO1nxks8zZgsIXV5Ri_49hB2jFz14n7LXRUjV9g0O0rN-r0Wx6FPbPB3wowZQ
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3Nb9MwFH-CggQc-BibKAzwgcsO3mI7du3jKFT76KqJtttuUWI_axVSWrUZfz-2k_KhiQO3SIlj6z0n78Pv93sAn7KSG2SW0ZJXWSTV9lQ77akP9iGE2EZVCSF3NR5MJvrmxlx2YPWEhUHEVHyGh_EyneW7pb2LqbIj1vKhqYfwSOY5z1q41q-USqQW4Smjx5niNEw22B5MZuZofD0M4SCXh0LGLnTsL0OUOqvc-x0nGzN68Z-rewnPO2eSHLfafwUPsN6BZ39QDL6G22-4SXgrMozLoMdN0xY4ktnWZcU1_RxsmSMJjEvnYVOS4fSUjIJhq0r7nVwvmlvyBXFFzpaLuiHTlPCnEZhQY3jxMtq_XZiPvs6GJ7TrrkAtzwcN1VUpvEHtWM4cd9xIWQn0QVFS5R5F6ay1zA4qIbxjJs-8NLm3GBwSJsNtsQe9elnjGyCKW6W0c0ZG2JzRFcfMCK2MtiIWgfThYCvqYtWSaBQp-MhMEdRSRLUUnVr6sBtF-_u5Tqp92N8qp-g-s00RvEEV6RW1evuPYR_hycnsYlyMTyfn7-BpnKkt39uHXrO-w_fw2P5oFpv1h7SXfgK6gsTb
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=Residual+Cross-Attention+Transformer-Based+Multi-User+CSI+Feedback+With+Deep+Joint+Source-Channel+Coding&rft.jtitle=IEEE+wireless+communications+letters&rft.au=Zhang%2C+Hengwei&rft.au=Wu%2C+Minghui&rft.au=Qiao%2C+Li&rft.au=Liu%2C+Ling&rft.date=2025-08-01&rft.issn=2162-2337&rft.eissn=2162-2345&rft.volume=14&rft.issue=8&rft.spage=2481&rft.epage=2485&rft_id=info:doi/10.1109%2FLWC.2025.3574011&rft.externalDBID=n%2Fa&rft.externalDocID=10_1109_LWC_2025_3574011
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=2162-2337&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=2162-2337&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=2162-2337&client=summon