Slashing Communication Traffic in Federated Learning by Transmitting Clustered Model Updates
Federated Learning (FL) is an emerging decentralized learning framework through which multiple clients can collaboratively train a learning model. However, a major obstacle that impedes the wide deployment of FL lies in massive communication traffic. To train high dimensional machine learning models...
Uložené v:
| Vydané v: | IEEE journal on selected areas in communications Ročník 39; číslo 8; s. 2572 - 2589 |
|---|---|
| Hlavní autori: | , , , |
| Médium: | Journal Article |
| Jazyk: | English |
| Vydavateľské údaje: |
New York
IEEE
01.08.2021
The Institute of Electrical and Electronics Engineers, Inc. (IEEE) |
| Predmet: | |
| ISSN: | 0733-8716, 1558-0008 |
| On-line prístup: | Získať plný text |
| Tagy: |
Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
|
| Abstract | Federated Learning (FL) is an emerging decentralized learning framework through which multiple clients can collaboratively train a learning model. However, a major obstacle that impedes the wide deployment of FL lies in massive communication traffic. To train high dimensional machine learning models (such as CNN models), heavy communication traffic can be incurred by exchanging model updates via the Internet between clients and the parameter server (PS), implying that the network resource can be easily exhausted. Compressing model updates is an effective way to reduce the traffic amount. However, a flexible unbiased compression algorithm applicable for both uplink and downlink compression in FL is still absent from existing works. In this work, we devise the Model Update Compression by Soft Clustering (MUCSC) algorithm to compress model updates transmitted between clients and the PS. In MUCSC, it is only necessary to transmit cluster centroids and the cluster ID of each model update. Moreover, we prove that: 1) The compressed model updates are unbiased estimation of their original values so that the convergence rate by transmitting compressed model updates is unchanged; 2) MUCSC can guarantee that the influence of the compression error on the model accuracy is minimized. Then, we further propose the boosted MUCSC (B-MUCSC) algorithm, a biased compression algorithm that can achieve an extremely high compression rate by grouping insignificant model updates into a super cluster. B-MUCSC is suitable for scenarios with very scarce network resource. Ultimately, we conduct extensive experiments with the CIFAR-10 and FEMNIST datasets to demonstrate that our algorithms can not only substantially reduce the volume of communication traffic in FL, but also improve the training efficiency in practical networks. |
|---|---|
| AbstractList | Federated Learning (FL) is an emerging decentralized learning framework through which multiple clients can collaboratively train a learning model. However, a major obstacle that impedes the wide deployment of FL lies in massive communication traffic. To train high dimensional machine learning models (such as CNN models), heavy communication traffic can be incurred by exchanging model updates via the Internet between clients and the parameter server (PS), implying that the network resource can be easily exhausted. Compressing model updates is an effective way to reduce the traffic amount. However, a flexible unbiased compression algorithm applicable for both uplink and downlink compression in FL is still absent from existing works. In this work, we devise the Model Update Compression by Soft Clustering (MUCSC) algorithm to compress model updates transmitted between clients and the PS. In MUCSC, it is only necessary to transmit cluster centroids and the cluster ID of each model update. Moreover, we prove that: 1) The compressed model updates are unbiased estimation of their original values so that the convergence rate by transmitting compressed model updates is unchanged; 2) MUCSC can guarantee that the influence of the compression error on the model accuracy is minimized. Then, we further propose the boosted MUCSC (B-MUCSC) algorithm, a biased compression algorithm that can achieve an extremely high compression rate by grouping insignificant model updates into a super cluster. B-MUCSC is suitable for scenarios with very scarce network resource. Ultimately, we conduct extensive experiments with the CIFAR-10 and FEMNIST datasets to demonstrate that our algorithms can not only substantially reduce the volume of communication traffic in FL, but also improve the training efficiency in practical networks. |
| Author | Cui, Laizhong Pan, Yi Su, Xiaoxin Zhou, Yipeng |
| Author_xml | – sequence: 1 givenname: Laizhong orcidid: 0000-0003-1991-290X surname: Cui fullname: Cui, Laizhong email: cuilz@szu.edu.cn organization: College of Computer Science and Software Engineering, Shenzhen University, Shenzhen, China – sequence: 2 givenname: Xiaoxin orcidid: 0000-0001-9514-1102 surname: Su fullname: Su, Xiaoxin email: suxiaoxin2016@163.com organization: College of Computer Science and Software Engineering, Shenzhen University, Shenzhen, China – sequence: 3 givenname: Yipeng orcidid: 0000-0003-1533-0865 surname: Zhou fullname: Zhou, Yipeng email: yipeng.zhou@mq.edu.au organization: Department of Computing, Faculty of Science and Engineering, Macquarie University, Sydney, NSW, Australia – sequence: 4 givenname: Yi orcidid: 0000-0002-2766-3096 surname: Pan fullname: Pan, Yi email: yi.pan@siat.ac.cn organization: Faculty of Computer Science and Control Engineering, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China |
| BookMark | eNp9kL1OwzAURi1UJNrCAyCWSMwp13acOGNVUX5UxNB2Q7Kc2AFXiVNsZ-jbk7QVAwPTXc65n3QmaGRbqxG6xTDDGPKH1_V8MSNA8IwCz0hKLtAYM8ZjAOAjNIaM0phnOL1CE-93ADhJOBmjj3Ut_Zexn9GibZrOmlIG09po42RVmTIyNlpqpZ0MWkUrLZ0d2OIwANY3JoSjW3c-aNcjb63SdbTdq17w1-iykrXXN-c7Rdvl42bxHK_en14W81VckpyGWJaEKFLmjBVJTvM80bzQpSqYrnBKZMYVUSpLeSFpxTlNK0g4MAV5ArzAvKBTdH_6u3ftd6d9ELu2c7afFIQxglMAlvYUPlGla713uhJ7ZxrpDgKDGCKKIaIYIopzxN7J_jilCcdCwUlT_2venUyjtf5dyvvqmGH6A2Z_gUM |
| CODEN | ISACEM |
| CitedBy_id | crossref_primary_10_1016_j_sysarc_2023_102927 crossref_primary_10_1109_TBDATA_2024_3404104 crossref_primary_10_1109_JSAC_2023_3310102 crossref_primary_10_1109_TII_2022_3161517 crossref_primary_10_1016_j_comnet_2025_111233 crossref_primary_10_1007_s11432_021_3532_8 crossref_primary_10_1109_JSAC_2022_3213345 crossref_primary_10_1109_TPDS_2023_3265588 crossref_primary_10_1109_JSTSP_2024_3381373 crossref_primary_10_1109_TC_2024_3477971 crossref_primary_10_1109_TNET_2024_3363916 crossref_primary_10_1016_j_comnet_2024_110886 crossref_primary_10_1109_TCAD_2023_3307459 crossref_primary_10_1007_s13369_025_10533_y crossref_primary_10_1016_j_comnet_2024_110883 crossref_primary_10_1109_TON_2024_3520530 crossref_primary_10_1088_3049_477X_add26a crossref_primary_10_1109_TNNLS_2023_3294295 crossref_primary_10_1109_TNET_2022_3168939 crossref_primary_10_1109_JSAC_2024_3431568 crossref_primary_10_1109_TPDS_2023_3240883 crossref_primary_10_1109_TPDS_2024_3447221 crossref_primary_10_1109_TBDATA_2022_3222971 crossref_primary_10_1109_TDSC_2025_3559108 crossref_primary_10_1007_s10586_025_05310_3 crossref_primary_10_1109_TMC_2024_3504284 crossref_primary_10_1016_j_jfranklin_2022_12_053 |
| Cites_doi | 10.1109/GLOCOM.2018.8647927 10.1016/j.ijmedinf.2018.01.007 10.1109/TNNLS.2019.2944481 10.1109/MC.2017.3641638 10.1109/ICC40277.2020.9148987 10.1109/ICC40277.2020.9148862 10.1109/ICDCS.2019.00099 10.1145/3298981 10.1109/ICC.2019.8761315 10.1109/JSAC.2014.2328154 10.1109/VCIP.2018.8698609 10.1109/5.726791 10.24963/ijcai.2019/473 10.1109/INFOCOM.2019.8737464 10.3389/fams.2018.00062 10.1109/GLOCOM.2018.8647616 10.1109/GLOBECOM38437.2019.9013160 10.1109/COMST.2020.2986024 10.1609/aaai.v33i01.33015693 10.1109/ICC40277.2020.9148586 10.1109/JSAC.2019.2904348 |
| ContentType | Journal Article |
| Copyright | Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2021 |
| Copyright_xml | – notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2021 |
| DBID | 97E RIA RIE AAYXX CITATION 7SP 8FD L7M |
| DOI | 10.1109/JSAC.2021.3087262 |
| DatabaseName | IEEE All-Society Periodicals Package (ASPP) 2005–Present 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 | 1558-0008 |
| EndPage | 2589 |
| ExternalDocumentID | 10_1109_JSAC_2021_3087262 9448151 |
| Genre | orig-research |
| GrantInformation_xml | – fundername: Shenzhen Science and Technology Program grantid: RCYX20200714114645048; JCYJ20190808142207420; GJHZ20190822095416463 – fundername: National Key Research and Development Plan of China grantid: 2018YFB1800302; 2018YFB1800805 funderid: 10.13039/501100012166 – fundername: Project of ”FANet: PCL Future Greater-Bay Area Network Facilities for Large-scale Experiments and Applications grantid: LZC0019 – fundername: Australia Research Council grantid: DE180100950 funderid: 10.13039/501100000923 – fundername: National Natural Science Foundation of China grantid: 61772345 funderid: 10.13039/501100001809 – fundername: Pearl River Young Scholars funding of Shenzhen University funderid: 10.13039/501100009019 |
| GroupedDBID | -~X .DC 0R~ 29I 3EH 4.4 41~ 5GY 5VS 6IK 97E AAJGR AARMG AASAJ AAWTH ABAZT ABQJQ ABVLG ACGFO ACGFS ACIWK ACNCT ADRHT AENEX AETIX AGQYO AGSQL AHBIQ AI. AIBXA AKJIK AKQYR ALLEH ALMA_UNASSIGNED_HOLDINGS ATWAV BEFXN BFFAM BGNUA BKEBE BPEOZ CS3 DU5 EBS EJD HZ~ H~9 IBMZZ ICLAB IES IFIPE IFJZH IPLJI JAVBF LAI M43 O9- OCL P2P RIA RIE RNS TN5 VH1 AAYXX CITATION 7SP 8FD L7M |
| ID | FETCH-LOGICAL-c293t-ac22d2c955b493994e8becdb5ef162a78d2dd768ba3f8836f04805d09408b18b3 |
| IEDL.DBID | RIE |
| ISICitedReferencesCount | 39 |
| ISICitedReferencesURI | http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000673624000023&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| ISSN | 0733-8716 |
| IngestDate | Sun Sep 07 07:11:14 EDT 2025 Sat Nov 29 03:23:03 EST 2025 Tue Nov 18 22:12:13 EST 2025 Wed Aug 27 02:40:51 EDT 2025 |
| IsPeerReviewed | true |
| IsScholarly | true |
| Issue | 8 |
| Language | English |
| LinkModel | DirectLink |
| MergedId | FETCHMERGED-LOGICAL-c293t-ac22d2c955b493994e8becdb5ef162a78d2dd768ba3f8836f04805d09408b18b3 |
| Notes | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14 |
| ORCID | 0000-0003-1991-290X 0000-0002-2766-3096 0000-0001-9514-1102 0000-0003-1533-0865 |
| PQID | 2552160056 |
| PQPubID | 85481 |
| PageCount | 18 |
| ParticipantIDs | ieee_primary_9448151 crossref_primary_10_1109_JSAC_2021_3087262 proquest_journals_2552160056 crossref_citationtrail_10_1109_JSAC_2021_3087262 |
| PublicationCentury | 2000 |
| PublicationDate | 2021-08-01 |
| PublicationDateYYYYMMDD | 2021-08-01 |
| PublicationDate_xml | – month: 08 year: 2021 text: 2021-08-01 day: 01 |
| PublicationDecade | 2020 |
| PublicationPlace | New York |
| PublicationPlace_xml | – name: New York |
| PublicationTitle | IEEE journal on selected areas in communications |
| PublicationTitleAbbrev | J-SAC |
| PublicationYear | 2021 |
| 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 mcmahan (ref17) 2017; 54 ref14 li (ref18) 2014 ref10 fikri aji (ref45) 2017 ref19 caldas (ref47) 2018 tobias springenberg (ref9) 2014 mishchenko (ref40) 2019 reinsel (ref1) 2018 ref51 kone?ný (ref15) 2018; 4 li (ref8) 2014 alistarh (ref32) 2017 ref48 alistarh (ref12) 2016 stich (ref35) 2018 ref41 hard (ref22) 2018 suresh (ref33) 2017 dinh (ref42) 2019 verma (ref21) 2018 ref49 ref4 ref3 krizhevsky (ref46) 2009 ref6 ref5 kone?ný (ref16) 2016 mcmahan (ref2) 2017 alistarh (ref38) 2018 ref37 bernstein (ref50) 2018 li (ref23) 2019 lin (ref44) 2017 ref24 ref26 ref25 bonawitz (ref7) 2019 ref20 (ref43) 2019 han (ref36) 2020 ref28 wang (ref34) 2018 ref27 ref29 shi (ref39) 2019 wen (ref30) 2017 kone?ný (ref11) 2016 wangni (ref31) 2018 |
| References_xml | – ident: ref6 doi: 10.1109/GLOCOM.2018.8647927 – year: 2016 ident: ref12 article-title: QSGD: Communication-efficient SGD via gradient quantization and encoding publication-title: arXiv 1610 02132 – year: 2014 ident: ref9 article-title: Striving for simplicity: The all convolutional net publication-title: arXiv 1412 6806 – ident: ref20 doi: 10.1016/j.ijmedinf.2018.01.007 – year: 2019 ident: ref43 publication-title: Speedtest Market Report – ident: ref13 doi: 10.1109/TNNLS.2019.2944481 – start-page: 1509 year: 2017 ident: ref30 article-title: Terngrad: Ternary gradients to reduce communication in distributed deep learning publication-title: Proc Adv Neural Inf Process Syst – ident: ref10 doi: 10.1109/MC.2017.3641638 – ident: ref14 doi: 10.1109/ICC40277.2020.9148987 – ident: ref29 doi: 10.1109/ICC40277.2020.9148862 – year: 2017 ident: ref44 article-title: Deep gradient compression: Reducing the communication bandwidth for distributed training publication-title: arXiv 1712 01887 – start-page: 583 year: 2014 ident: ref18 article-title: Scaling distributed machine learning with the parameter server publication-title: Proc Int Conf Big Data Sci Comput BigDataScience – start-page: 28 year: 2018 ident: ref1 publication-title: The Digitization of the World From Edge to Core – ident: ref37 doi: 10.1109/ICDCS.2019.00099 – ident: ref4 doi: 10.1145/3298981 – year: 2019 ident: ref39 article-title: Layer-wise adaptive gradient sparsification for distributed deep learning with convergence guarantees publication-title: arXiv 1911 08727 – start-page: 3329 year: 2017 ident: ref33 article-title: Distributed mean estimation with limited communication publication-title: Proc Int Conf Mach Learn – ident: ref3 doi: 10.1109/ICC.2019.8761315 – year: 2018 ident: ref21 article-title: Federated AI for building AI solutions across multiple agencies publication-title: arXiv 1809 10036 – start-page: 19 year: 2014 ident: ref8 article-title: Communication efficient distributed machine learning with the parameter server publication-title: Proc Adv Neural Inf Process Syst – start-page: 1299 year: 2018 ident: ref31 article-title: Gradient sparsification for communication-efficient distributed optimization publication-title: Proc Adv Neural Inf Process Syst – year: 2009 ident: ref46 article-title: Learning multiple layers of features from tiny images – year: 2016 ident: ref11 article-title: Federated learning: Strategies for improving communication efficiency publication-title: arXiv 1610 05492 – year: 2017 ident: ref2 publication-title: Federated learning Collaborative machine learning without centralized training data – year: 2020 ident: ref36 article-title: Adaptive gradient sparsification for efficient federated learning: An online learning approach publication-title: arXiv 2001 04756 – ident: ref49 doi: 10.1109/JSAC.2014.2328154 – year: 2017 ident: ref45 article-title: Sparse communication for distributed gradient descent publication-title: arXiv 1704 05021 – start-page: 1709 year: 2017 ident: ref32 article-title: QSGD: Communication-efficient SGD via gradient quantization and encoding publication-title: Proc Adv Neural Inf Process Syst – start-page: 4447 year: 2018 ident: ref35 article-title: Sparsified SGD with memory publication-title: Proc Adv Neural Inf Process Syst – ident: ref28 doi: 10.1109/VCIP.2018.8698609 – year: 2019 ident: ref23 article-title: On the convergence of FedAvg on non-IID data publication-title: arXiv 1907 02189 – year: 2018 ident: ref47 article-title: LEAF: A benchmark for federated settings publication-title: arXiv 1812 01097 – year: 2018 ident: ref22 article-title: Federated learning for mobile keyboard prediction publication-title: arXiv 1811 03604 – year: 2019 ident: ref42 article-title: Federated learning over wireless networks: Convergence analysis and resource allocation publication-title: arXiv 1910 13067 – ident: ref48 doi: 10.1109/5.726791 – ident: ref51 doi: 10.24963/ijcai.2019/473 – ident: ref25 doi: 10.1109/INFOCOM.2019.8737464 – start-page: 5973 year: 2018 ident: ref38 article-title: The convergence of sparsified gradient methods publication-title: Proc Adv Neural Inf Process Syst – volume: 4 start-page: 62 year: 2018 ident: ref15 article-title: Randomized distributed mean estimation: Accuracy vs. communication publication-title: Frontiers Appl Math Statist doi: 10.3389/fams.2018.00062 – year: 2019 ident: ref7 article-title: Towards federated learning at scale: System design publication-title: arXiv 1902 01046 – ident: ref5 doi: 10.1109/GLOCOM.2018.8647616 – ident: ref26 doi: 10.1109/GLOBECOM38437.2019.9013160 – year: 2016 ident: ref16 article-title: Federated optimization: Distributed machine learning for on-device intelligence publication-title: arXiv 1610 02527 – start-page: 9850 year: 2018 ident: ref34 article-title: ATOMO: Communication-efficient learning via atomic sparsification publication-title: Proc Adv Neural Inf Process Syst – ident: ref19 doi: 10.1109/COMST.2020.2986024 – year: 2019 ident: ref40 article-title: Distributed learning with compressed gradient differences publication-title: arXiv 1901 09269 – ident: ref41 doi: 10.1609/aaai.v33i01.33015693 – ident: ref27 doi: 10.1109/ICC40277.2020.9148586 – ident: ref24 doi: 10.1109/JSAC.2019.2904348 – volume: 54 start-page: 1273 year: 2017 ident: ref17 article-title: Communication-efficient learning of deep networks from decentralized data publication-title: Proc Int Conf Artif Intell Statist (AISTATS) – year: 2018 ident: ref50 article-title: SignSGD: Compressed optimisation for non-convex problems publication-title: arXiv 1802 04434 |
| SSID | ssj0014482 |
| Score | 2.553327 |
| Snippet | Federated Learning (FL) is an emerging decentralized learning framework through which multiple clients can collaboratively train a learning model. However, a... |
| SourceID | proquest crossref ieee |
| SourceType | Aggregation Database Enrichment Source Index Database Publisher |
| StartPage | 2572 |
| SubjectTerms | Adaptation models Algorithms Centroids Clients Clustering Communication Communications traffic Compressing Compression algorithms Computational modeling Convergence convergence rate Data models Federated learning Machine learning Model accuracy model update compression Traffic models Training Transmission Uplink |
| Title | Slashing Communication Traffic in Federated Learning by Transmitting Clustered Model Updates |
| URI | https://ieeexplore.ieee.org/document/9448151 https://www.proquest.com/docview/2552160056 |
| Volume | 39 |
| WOSCitedRecordID | wos000673624000023&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: 1558-0008 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0014482 issn: 0733-8716 databaseCode: RIE dateStart: 19830101 isFulltext: true titleUrlDefault: https://ieeexplore.ieee.org/ providerName: IEEE |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LSwMxEA61eNCDrypWq-TgSUzdzb6SYykWESlCrfQgLJvHSmHdlj4E_72ZbFosiuBtYSdhyZfNzGRmvkHoKvMET8wpSLiUjIS-jIjgPCdCJDIzFkoQ2qzKl8ek32ejEX-qoZt1LYzW2iaf6TY82li-msglXJXd8hCoRYyvs5UkcVWrtY4YmFc2YpAEAQEnwEUwfY_fPgw6XeMJUr8N9Hc0phs6yDZV-XESW_XS2__fhx2gPWdG4k6F-yGq6fII7X4jF2yg10FRNUrCG0Ug2GgnoI3A4xL3gEnCGJsKO5rVNyw-sVVf72ObEI27xRKoFIwIdE0r8HAKVwTzYzTs3T1374nrpUCkUegLkklKFZU8ikTIjVESambQUyLSuR_TLGGKKmVcD5EFOWNBnEOteaSAXY8Jn4ngBNXLSalPEQZG-1hJX-s4gGhwZuaguQippp7wItFE3mp1U-mIxqHfRZFah8PjKQCSAiCpA6SJrtdDphXLxl_CDUBgLegWv4laKwhT9x_OU-MwUT8GvtOz30edox2Yu0rpa6H6YrbUF2hbfizG89ml3WJfDSfN5g |
| linkProvider | IEEE |
| linkToHtml | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LS8NAEB6KCurBVxWrVffgSUybbJ57LMVStRahrfQghOwjUqht6UPw37uzSYtFEbwFMpuE_TY7Mzsz3wBcJzZnod4FLSZEZHmO8C3OWGpxHopEWyiuZ7IqX1phux31--y5ALerWhillEk-UxW8NLF8ORYLPCqrMg-pRbSvs4mds_ysWmsVM9A3TcwgdF0L3YA8hunYrPrQqdW1L0idChLg0YCuaSHTVuXHXmwUTGP_f592AHu5IUlqGfKHUFCjI9j9Ri9YhNfOMGuVRNbKQIjWT0gcQQYj0kAuCW1uSpITrb4R_kmMAnsfmJRoUh8ukExBi2DftCHpTfCQYHYMvcZdt9608m4KltAqfW4lglJJBfN97jFtlngq0vhJ7qvUCWgSRpJKqZ0PnrhpFLlBitXmvkR-vYg7EXdPYGM0HqlTIMhpH0jhKBW4GA9O9DNoyj2qqM1tn5fAXs5uLHKqcex4MYyNy2GzGAGJEZA4B6QEN6shk4xn4y_hIiKwEswnvwTlJYRx_ifOYu0yUSdAxtOz30ddwXaz-9SKW_ftx3PYwfdkCX5l2JhPF-oCtsTHfDCbXprl9gX2O9Ex |
| 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=Slashing+Communication+Traffic+in+Federated+Learning+by+Transmitting+Clustered+Model+Updates&rft.jtitle=IEEE+journal+on+selected+areas+in+communications&rft.au=Cui%2C+Laizhong&rft.au=Su%2C+Xiaoxin&rft.au=Zhou%2C+Yipeng&rft.au=Pan%2C+Yi&rft.date=2021-08-01&rft.pub=The+Institute+of+Electrical+and+Electronics+Engineers%2C+Inc.+%28IEEE%29&rft.issn=0733-8716&rft.eissn=1558-0008&rft.volume=39&rft.issue=8&rft.spage=2572&rft_id=info:doi/10.1109%2FJSAC.2021.3087262&rft.externalDBID=NO_FULL_TEXT |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0733-8716&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0733-8716&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0733-8716&client=summon |