Byzantine-Resilient Secure Federated Learning on Low-Bandwidth Networks
Privacy-preserving and Byzantine-resilient machine learning has been an important research issue, and many centralized methods have been developed. However, it is difficult for these methods to achieve fast learning and high accuracy simultaneously. In contrast, federated learning based on local mod...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10129854/ |
| _version_ | 1827934704718839808 |
|---|---|
| author | Hiroki Masuda Kentaro Kita Yuki Koizumi Junji Takemasa Toru Hasegawa |
| author_facet | Hiroki Masuda Kentaro Kita Yuki Koizumi Junji Takemasa Toru Hasegawa |
| author_sort | Hiroki Masuda |
| collection | DOAJ |
| description | Privacy-preserving and Byzantine-resilient machine learning has been an important research issue, and many centralized methods have been developed. However, it is difficult for these methods to achieve fast learning and high accuracy simultaneously. In contrast, federated learning based on local model masking like Byzantine-Resilient Secure Aggregation (BREA), is a promising approach to simultaneously achieve them. Despite the advantage of light computation of randomizing local models of users for privacy preservation, the verification of shares generated from local models in BREA, which mitigates Byzantine attacks, still incurs large complexity in communication. The paper designs a share verification method for BREA to offload some parts of the share verification process from users to a semi-honest server, which avoids broadcasting large-size commitments to shares. In addition, to mitigate the increase in computation time due to computations offloaded to the server, our method makes the verification algorithm running on the server efficient and executes the server and user computations in parallel. In our experiments, our method provides a speedup of up to <inline-formula> <tex-math notation="LaTeX">$15\times $ </tex-math></inline-formula> on low-bandwidth networks like mobile networks. Our method also preserves BREA’s resilience against Byzantine attacks. |
| first_indexed | 2024-03-13T07:44:33Z |
| format | Article |
| id | doaj.art-1dcd1cf01c4640b7be7b9405e4a85901 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-03-13T07:44:33Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-1dcd1cf01c4640b7be7b9405e4a859012023-06-02T23:00:21ZengIEEEIEEE Access2169-35362023-01-0111517545176610.1109/ACCESS.2023.327785810129854Byzantine-Resilient Secure Federated Learning on Low-Bandwidth NetworksHiroki Masuda0https://orcid.org/0009-0002-3948-3671Kentaro Kita1https://orcid.org/0000-0002-7982-3530Yuki Koizumi2https://orcid.org/0000-0002-9254-6558Junji Takemasa3https://orcid.org/0000-0002-5361-1855Toru Hasegawa4https://orcid.org/0000-0002-8925-1732Graduate School of Information Science and Technology, Osaka University, Suita, Osaka, JapanGraduate School of Information Science and Technology, Osaka University, Suita, Osaka, JapanGraduate School of Information Science and Technology, Osaka University, Suita, Osaka, JapanGraduate School of Information Science and Technology, Osaka University, Suita, Osaka, JapanGraduate School of Information Science and Technology, Osaka University, Suita, Osaka, JapanPrivacy-preserving and Byzantine-resilient machine learning has been an important research issue, and many centralized methods have been developed. However, it is difficult for these methods to achieve fast learning and high accuracy simultaneously. In contrast, federated learning based on local model masking like Byzantine-Resilient Secure Aggregation (BREA), is a promising approach to simultaneously achieve them. Despite the advantage of light computation of randomizing local models of users for privacy preservation, the verification of shares generated from local models in BREA, which mitigates Byzantine attacks, still incurs large complexity in communication. The paper designs a share verification method for BREA to offload some parts of the share verification process from users to a semi-honest server, which avoids broadcasting large-size commitments to shares. In addition, to mitigate the increase in computation time due to computations offloaded to the server, our method makes the verification algorithm running on the server efficient and executes the server and user computations in parallel. In our experiments, our method provides a speedup of up to <inline-formula> <tex-math notation="LaTeX">$15\times $ </tex-math></inline-formula> on low-bandwidth networks like mobile networks. Our method also preserves BREA’s resilience against Byzantine attacks.https://ieeexplore.ieee.org/document/10129854/Federated learningprivacysecurity |
| spellingShingle | Hiroki Masuda Kentaro Kita Yuki Koizumi Junji Takemasa Toru Hasegawa Byzantine-Resilient Secure Federated Learning on Low-Bandwidth Networks IEEE Access Federated learning privacy security |
| title | Byzantine-Resilient Secure Federated Learning on Low-Bandwidth Networks |
| title_full | Byzantine-Resilient Secure Federated Learning on Low-Bandwidth Networks |
| title_fullStr | Byzantine-Resilient Secure Federated Learning on Low-Bandwidth Networks |
| title_full_unstemmed | Byzantine-Resilient Secure Federated Learning on Low-Bandwidth Networks |
| title_short | Byzantine-Resilient Secure Federated Learning on Low-Bandwidth Networks |
| title_sort | byzantine resilient secure federated learning on low bandwidth networks |
| topic | Federated learning privacy security |
| url | https://ieeexplore.ieee.org/document/10129854/ |
| work_keys_str_mv | AT hirokimasuda byzantineresilientsecurefederatedlearningonlowbandwidthnetworks AT kentarokita byzantineresilientsecurefederatedlearningonlowbandwidthnetworks AT yukikoizumi byzantineresilientsecurefederatedlearningonlowbandwidthnetworks AT junjitakemasa byzantineresilientsecurefederatedlearningonlowbandwidthnetworks AT toruhasegawa byzantineresilientsecurefederatedlearningonlowbandwidthnetworks |