Privacy-Preserving Federated Learning With Resource-Adaptive Compression for Edge Devices
Federated learning (FL) has gained widespread attention as a distributed machine learning (ML) technique that offers data protection when training on local devices. Unlike conventional centralized training in traditional ML, FL incorporates privacy and security measures as it does not share raw data...
Saved in:
| Published in: | IEEE internet of things journal Vol. 11; no. 8; pp. 13180 - 13198 |
|---|---|
| Main Authors: | , , |
| Format: | Journal Article |
| Language: | English |
| Published: |
Piscataway
IEEE
15.04.2024
The Institute of Electrical and Electronics Engineers, Inc. (IEEE) |
| Subjects: | |
| ISSN: | 2327-4662, 2327-4662 |
| Online Access: | Get full text |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Abstract | Federated learning (FL) has gained widespread attention as a distributed machine learning (ML) technique that offers data protection when training on local devices. Unlike conventional centralized training in traditional ML, FL incorporates privacy and security measures as it does not share raw data between the client and server, thereby safeguarding potentially sensitive information. However, there are still vulnerabilities in the FL field, and commonly used approaches, such as encryption and blockchain technologies, often result in significant computational and communication costs, making them impractical for devices with restricted resources. To tackle this challenge, we present a privacy-preserving FL system specifically designed for resource-constrained devices, leveraging compressive sensing and differential privacy (DP) techniques. We implemented the weight-pruning-based compressive sensing method with an adaptive compression ratio based on resource availability. In addition, we employ DP to introduce noise to the gradient before sending it to a central server for aggregation, thereby protecting the gradient's privacy. Evaluation results demonstrate that our proposed method achieves slightly better accuracy when compared to state-of-the-art methods like DP-federated averaging, DP-FedOpt, and adaptive Gaussian clipping-DP (AGC-DP) for the MNIST, Fashion-MNIST, and Human Activity Recognition data sets. Furthermore, our approach achieves this higher accuracy with a lower total communication cost and training time than the current state-of-the-art methods. Moreover, we comprehensively evaluate our method's resilience against poisoning attacks, revealing its better resistance than existing state-of-the-art approaches. |
|---|---|
| AbstractList | Federated learning (FL) has gained widespread attention as a distributed machine learning (ML) technique that offers data protection when training on local devices. Unlike conventional centralized training in traditional ML, FL incorporates privacy and security measures as it does not share raw data between the client and server, thereby safeguarding potentially sensitive information. However, there are still vulnerabilities in the FL field, and commonly used approaches, such as encryption and blockchain technologies, often result in significant computational and communication costs, making them impractical for devices with restricted resources. To tackle this challenge, we present a privacy-preserving FL system specifically designed for resource-constrained devices, leveraging compressive sensing and differential privacy (DP) techniques. We implemented the weight-pruning-based compressive sensing method with an adaptive compression ratio based on resource availability. In addition, we employ DP to introduce noise to the gradient before sending it to a central server for aggregation, thereby protecting the gradient’s privacy. Evaluation results demonstrate that our proposed method achieves slightly better accuracy when compared to state-of-the-art methods like DP-federated averaging, DP-FedOpt, and adaptive Gaussian clipping-DP (AGC-DP) for the MNIST, Fashion-MNIST, and Human Activity Recognition data sets. Furthermore, our approach achieves this higher accuracy with a lower total communication cost and training time than the current state-of-the-art methods. Moreover, we comprehensively evaluate our method’s resilience against poisoning attacks, revealing its better resistance than existing state-of-the-art approaches. |
| Author | Hidayat, Muhammad Ayat Nakamura, Yugo Arakawa, Yutaka |
| Author_xml | – sequence: 1 givenname: Muhammad Ayat orcidid: 0000-0002-1841-7010 surname: Hidayat fullname: Hidayat, Muhammad Ayat email: muhammad.971@s.kyushu-u.ac.jp organization: Department of Information Science and Technology, Graduate School of Information Science and Electrical Engineering, Kyushu University, Fukuoka, Japan – sequence: 2 givenname: Yugo orcidid: 0000-0002-8834-5323 surname: Nakamura fullname: Nakamura, Yugo email: y-nakamura@ait.kyushu-u.ac.jp organization: Department of Information Science and Technology, Graduate School of Information Science and Electrical Engineering, Kyushu University, Fukuoka, Japan – sequence: 3 givenname: Yutaka orcidid: 0000-0002-7156-9160 surname: Arakawa fullname: Arakawa, Yutaka email: arakawa@ait.kyushu-u.ac.jp organization: Department of Information Science and Technology, Graduate School of Information Science and Electrical Engineering, Kyushu University, Fukuoka, Japan |
| BookMark | eNp9kEtrAjEQgEOxUGv9AYUeFnpem8du4h7FamsRlGIpPS3ZZGIjumuTuOC_7y56kB56mmGYbx7fLeqUVQkI3RM8IARnT2-zxWpAMWUDxhKRpvQKdSmjIk44p52L_Ab1vd9gjBssJRnvoq-ls7VUx3jpwIOrbbmOpqDByQA6moN0ZVv6tOE7egdfHZyCeKTlPtgaonG12zect1UZmcpFE72G6Blqq8DfoWsjtx7659hDH9PJavwazxcvs_FoHiuaJSEWmolCckOHaogVodzIojAJGCy41glPM84UI0BNKrk0Q2MyzTXgwghpikKzHno8zd276ucAPuSb5sqyWZkzzFIqGiOk6SKnLuUq7x2YfO_sTrpjTnDeSsxbiXkrMT9LbBjxh1E2yNA8G5y023_JhxNpAeBiExMpTQj7BYM5gsI |
| CODEN | IITJAU |
| CitedBy_id | crossref_primary_10_1007_s10722_024_02277_9 crossref_primary_10_3390_healthcare12242587 crossref_primary_10_1109_SR_2025_3548547 crossref_primary_10_1109_LSENS_2024_3422417 crossref_primary_10_1109_JIOT_2025_3533003 crossref_primary_10_1109_JIOT_2024_3443642 crossref_primary_10_1016_j_sigpro_2025_110206 crossref_primary_10_1109_ACCESS_2024_3418016 |
| Cites_doi | 10.3390/app12199901 10.1109/ICEIC57457.2023.10049966 10.3389/frcmn.2021.657653 10.3390/electronics12020260 10.1145/3564625.3567973 10.1109/JIOT.2023.3267112 10.1109/TNSE.2021.3100096 10.1109/TrustCom56396.2022.00094 10.1109/MIS.2020.3010335 10.1109/TNNLS.2019.2944481 10.1109/SP.2019.00029 10.1109/TrustCom/BigDataSE.2019.00057 10.1007/s00500-021-06496-5 10.1109/ICDE53745.2022.00062 10.1109/MIS.2021.3082561 10.1145/2976749.2978318 10.1109/JIOT.2022.3171767 10.1109/TKDE.2021.3124599 10.1109/JIOT.2022.3189361 10.1109/JIOT.2022.3201117 10.1016/j.future.2020.10.007 10.1109/TNNLS.2022.3166101 10.3390/e24111545 10.1109/CCCI52664.2021.9583220 10.1109/IJCNN55064.2022.9889795 10.1109/TIFS.2020.2988575 10.1109/TCOMM.2021.3124961 10.1109/JIOT.2021.3127886 10.1109/TNSE.2022.3185327 10.1109/OJCS.2021.3099108 10.1109/mdm58254.2023.00042 10.1109/TII.2022.3195896 10.1109/SCISISIS55246.2022.10001931 10.1109/ICTC55196.2022.9952435 10.1109/GLOBECOM48099.2022.10001481 10.1109/TC.2023.3239542 10.1109/DSA52907.2021.00081 10.1145/3589462.3589502 |
| ContentType | Journal Article |
| Copyright | Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2024 |
| Copyright_xml | – notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2024 |
| DBID | 97E RIA RIE AAYXX CITATION 7SC 8FD JQ2 L7M L~C L~D |
| DOI | 10.1109/JIOT.2023.3347552 |
| DatabaseName | IEEE Xplore (IEEE) 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 Xplore url: https://ieeexplore.ieee.org/ sourceTypes: Publisher |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Computer Science |
| EISSN | 2327-4662 |
| EndPage | 13198 |
| ExternalDocumentID | 10_1109_JIOT_2023_3347552 10375241 |
| Genre | orig-research |
| GrantInformation_xml | – fundername: Ministry of Education, Culture, Sports, Science and Technology of Japan (MEXT) – fundername: Japan Science and Technology Agency (JST), PRESTO grantid: JPMJPR21P7 – fundername: Japan Society for the Promotion of Science (JSPS) KAKENHI grantid: JP19KT0020 funderid: 10.13039/501100001691 |
| 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 AAYXX CITATION 7SC 8FD JQ2 L7M L~C L~D |
| ID | FETCH-LOGICAL-c294t-7d37ba6f28c80c126fabbf4ef076dd465963c31e2f5a6af8ff9d6de0bf7afbbd3 |
| IEDL.DBID | RIE |
| ISICitedReferencesCount | 9 |
| ISICitedReferencesURI | http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001203466500130&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 Nov 03 04:24:43 EST 2025 Tue Nov 18 19:37:54 EST 2025 Sat Nov 29 01:43:58 EST 2025 Wed Aug 27 02:09:30 EDT 2025 |
| IsPeerReviewed | false |
| 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-c294t-7d37ba6f28c80c126fabbf4ef076dd465963c31e2f5a6af8ff9d6de0bf7afbbd3 |
| Notes | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14 |
| ORCID | 0000-0002-8834-5323 0000-0002-1841-7010 0000-0002-7156-9160 |
| PQID | 3035274751 |
| PQPubID | 2040421 |
| PageCount | 19 |
| ParticipantIDs | ieee_primary_10375241 proquest_journals_3035274751 crossref_primary_10_1109_JIOT_2023_3347552 crossref_citationtrail_10_1109_JIOT_2023_3347552 |
| PublicationCentury | 2000 |
| PublicationDate | 2024-04-15 |
| PublicationDateYYYYMMDD | 2024-04-15 |
| PublicationDate_xml | – month: 04 year: 2024 text: 2024-04-15 day: 15 |
| PublicationDecade | 2020 |
| PublicationPlace | Piscataway |
| PublicationPlace_xml | – name: Piscataway |
| PublicationTitle | IEEE internet of things journal |
| PublicationTitleAbbrev | JIoT |
| PublicationYear | 2024 |
| 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 Zari (ref17) 2021 ref30 ref33 ref10 ref32 ref2 ref1 ref39 ref16 ref38 Sikandar (ref8) 2023; 12 ref19 ref18 Li (ref31) 2021; 2021 Hu (ref23) 2021; 2 ref24 ref26 ref20 ref42 ref41 ref22 Mothukuri (ref11) 2021; 115 ref21 ref28 ref27 ref29 ref7 Kurniawan (ref9) 2022; 24 ref4 ref3 ref6 ref5 Reddi (ref40) 2020 Wang (ref25) |
| References_xml | – ident: ref1 doi: 10.3390/app12199901 – ident: ref4 doi: 10.1109/ICEIC57457.2023.10049966 – ident: ref29 doi: 10.3389/frcmn.2021.657653 – volume: 12 start-page: 260 issue: 2 year: 2023 ident: ref8 article-title: A detailed survey on federated learning attacks and defenses publication-title: Electronics doi: 10.3390/electronics12020260 – ident: ref26 doi: 10.1145/3564625.3567973 – ident: ref28 doi: 10.1109/JIOT.2023.3267112 – ident: ref35 doi: 10.1109/TNSE.2021.3100096 – ident: ref24 doi: 10.1109/TrustCom56396.2022.00094 – ident: ref37 doi: 10.1109/MIS.2020.3010335 – ident: ref14 doi: 10.1109/TNNLS.2019.2944481 – ident: ref16 doi: 10.1109/SP.2019.00029 – year: 2020 ident: ref40 article-title: Adaptive federated optimization publication-title: arXiv:2003.00295 – ident: ref18 doi: 10.1109/TrustCom/BigDataSE.2019.00057 – ident: ref10 doi: 10.1007/s00500-021-06496-5 – ident: ref32 doi: 10.1109/ICDE53745.2022.00062 – ident: ref7 doi: 10.1109/MIS.2021.3082561 – year: 2021 ident: ref17 article-title: Efficient passive membership inference attack in federated learning publication-title: arXiv:2111.00430 – ident: ref22 doi: 10.1145/2976749.2978318 – ident: ref36 doi: 10.1109/JIOT.2022.3171767 – ident: ref3 doi: 10.1109/TKDE.2021.3124599 – ident: ref41 doi: 10.1109/JIOT.2022.3189361 – ident: ref21 doi: 10.1109/JIOT.2022.3201117 – volume: 2021 start-page: 1 year: 2021 ident: ref31 article-title: An adaptive communication-efficient federated learning to resist gradient-based reconstruction attacks publication-title: Security Commun. Netw. – volume: 115 start-page: 619 year: 2021 ident: ref11 article-title: A survey on security and privacy of federated learning publication-title: Future Gener. Comput. Syst. doi: 10.1016/j.future.2020.10.007 – ident: ref33 doi: 10.1109/TNNLS.2022.3166101 – volume: 24 start-page: 1545 issue: 11 year: 2022 ident: ref9 article-title: Homomorphic encryption-based federated privacy preservation for deep active learning publication-title: Entropy doi: 10.3390/e24111545 – ident: ref39 doi: 10.1109/CCCI52664.2021.9583220 – ident: ref38 doi: 10.1109/IJCNN55064.2022.9889795 – ident: ref12 doi: 10.1109/TIFS.2020.2988575 – ident: ref15 doi: 10.1109/TCOMM.2021.3124961 – ident: ref30 doi: 10.1109/JIOT.2021.3127886 – ident: ref20 doi: 10.1109/TNSE.2022.3185327 – volume: 2 start-page: 276 year: 2021 ident: ref23 article-title: Concentrated differentially private federated learning with performance analysis publication-title: IEEE Open J. Comput. Soc. doi: 10.1109/OJCS.2021.3099108 – start-page: 22802 volume-title: Proc. Int. Conf. Mach. Learn. ident: ref25 article-title: Communication-efficient adaptive federated learning – ident: ref13 doi: 10.1109/mdm58254.2023.00042 – ident: ref2 doi: 10.1109/TII.2022.3195896 – ident: ref5 doi: 10.1109/SCISISIS55246.2022.10001931 – ident: ref6 doi: 10.1109/ICTC55196.2022.9952435 – ident: ref27 doi: 10.1109/GLOBECOM48099.2022.10001481 – ident: ref34 doi: 10.1109/TC.2023.3239542 – ident: ref42 doi: 10.1109/DSA52907.2021.00081 – ident: ref19 doi: 10.1145/3589462.3589502 |
| SSID | ssj0001105196 |
| Score | 2.3595977 |
| Snippet | Federated learning (FL) has gained widespread attention as a distributed machine learning (ML) technique that offers data protection when training on local... |
| SourceID | proquest crossref ieee |
| SourceType | Aggregation Database Enrichment Source Index Database Publisher |
| StartPage | 13180 |
| SubjectTerms | Adaptation models Blockchain Compression ratio Compressive sensing Computational modeling Cryptography Devices Differential privacy Federated learning federated learning (FL) Human activity recognition Machine learning Privacy privacy-preserving resource-adaptive Security Servers State of the art weight-pruning |
| Title | Privacy-Preserving Federated Learning With Resource-Adaptive Compression for Edge Devices |
| URI | https://ieeexplore.ieee.org/document/10375241 https://www.proquest.com/docview/3035274751 |
| Volume | 11 |
| WOSCitedRecordID | wos001203466500130&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 Xplore 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/eLvHCXMwlV1LS8QwEA4qHry4PnF9kYMnIWvbPHsUdVEP6kFRTyXPdUGqrOuC_95OmvWBKHgpPUxCydeZTCYz3yC0p43IaEEZKbRihAXNiS6ZIIbq0ofGYZBWxWYT8uJC3d2VV6lYPdbCeO9j8pnvwWu8y3dP9hVCZQdQ08YLKFOflVK0xVqfAZUcvBGRbi7zrDw4P7u87kF78B6lTHJefNt7YjOVHxY4biv9zj8_aAktJv8RH7aAL6MZX6-gzrQ3A06quorur0bDibZvBHIswB7UA9wH4ojGt3Q4saoO8O1w_ICnIXxy6PQzmD8M87UJsjVuvFp84gYeH_toVdbQTf_k-uiUpDYKxBYlGxPpqDRahEJZldm8EEEbE5gPmRTOMcEbHbQ090XgWuigQiidcD4zQepgjKPraK5-qv0GwkFK2cgIo1jJuAIywebBDAvcW2WzLsqmC1zZxDEOrS4eq3jWyMoKMKkAkyph0kX7H0OeW4KNv4TXAIQvgu36d9H2FMYq6eBLRYHqtTkt8Xzzl2FbaKGZPSbi5HwbzY1Hr34HzdvJePgy2o2_1zut9M9W |
| linkProvider | IEEE |
| linkToHtml | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LTxRBEO4QINGLiGJcQOmDJ5Neevo9R4JseLlyWCOeJv1cNzEDWRYS_j1TPb0qMZh4mcyhembS31R1dXXVVwh9sE5RzrggzBpBRLKS2Foo4ritY-ocBu1Nbjahx2NzeVlflGL1XAsTY8zJZ3EIt_ksP1z5WwiV7UNNm2RQpr4mhWC0L9f6HVKpwB9R5eyyovX-6cmXyRAahA85F1pK9mj1ye1U_rLBeWEZbfznJ71EL4oHiQ96yDfRSmxfoY1ldwZclPU1-n4xn91Zf08gywIsQjvFI6CO6LzLgAuv6hR_my1-4GUQnxwEew0GEMPz-hTZFnd-LT4K04g_xWxXttDX0dHk8JiURgrEs1osiA5cO6sSM95QXzGVrHNJxES1CkEo2Wmh51VkSVplk0mpDipE6pK2ybnA36DV9qqNbxFOWutORjkjaiEN0Al2F-FEktEbTweILie48YVlHJpd_GzyboPWDWDSACZNwWSAPv4act1TbPxLeAtA-EOwn_8B2l3C2BQtvGk4kL12-yVZbT8xbA89O558Pm_OT8ZnO-h596acllPJXbS6mN_Gd2jd3y1mN_P3-Vd7AJ0a0p0 |
| 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=Privacy-Preserving+Federated+Learning+With+Resource-Adaptive+Compression+for+Edge+Devices&rft.jtitle=IEEE+internet+of+things+journal&rft.au=Hidayat%2C+Muhammad+Ayat&rft.au=Nakamura%2C+Yugo&rft.au=Arakawa%2C+Yutaka&rft.date=2024-04-15&rft.issn=2327-4662&rft.eissn=2327-4662&rft.volume=11&rft.issue=8&rft.spage=13180&rft.epage=13198&rft_id=info:doi/10.1109%2FJIOT.2023.3347552&rft.externalDBID=n%2Fa&rft.externalDocID=10_1109_JIOT_2023_3347552 |
| 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 |