Practical Secure Aggregation for Privacy-Preserving Machine Learning
The DOI matches its Crossref record; 56 of the 56 references it cites were found in a registry or an index.
Abstract
We design a novel, communication-efficient, failure-robust protocol for secure aggregation of high-dimensional data. Our protocol allows a server to compute the sum of large, user-held data vectors from mobile devices in a secure manner (i.e. without learning each user's individual contribution), and can be used, for example, in a federated learning setting, to aggregate user-provided model updates for a deep neural network. We prove the security of our protocol in the honest-but-curious and active adversary settings, and show that security is maintained even if an arbitrarily chosen subset of users drop out at any time. We evaluate the efficiency of our protocol and show, by complexity analysis and a concrete implementation, that its runtime and communication overhead remain low even on large data sets and client pools. For 16-bit input values, our protocol offers $1.73 x communication expansion for 210 users and 220-dimensional vectors, and 1.98 x expansion for 214 users and 224-dimensional vectors over sending data in the clear.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Although SMC guarantees that none of the parties shares anything with each other or with any third party, it can not prevent an adversary from learning some individual information”
From Federated Learning for Healthcare Informatics (2021) · cites this paper as [8]
In the citing paper
“By combining local inference with trusted aggregation, TDAI can enhance both privacy and robustness of anomaly detection pipelines [239].”
From Anomaly detection and root-cause identification in microservices: a survey (2026) · cites this paper as [239]
In the citing paper
“FL is still in its infancy and is still vulnerable to various challenges [148,149].”
From Smart Wearables for the Detection of Cardiovascular Diseases: A Systematic Literature Review (2023) · cites this paper as [148]
In the citing paper
“Therefore, many FL frameworks use secure aggregation [81], [82], [83], [84] algorithms in the aggregation process, at which point most of the poisoning attack defense methods mentioned above will be invalid.”
From Poisoning attacks in federated learning: A survey (2023) · cites this paper as [81]
In the citing paper
“Bonawitz et al. [62] introduce a secure aggregation protocol for protecting individual model updates.”
From Handling privacy-sensitive medical data with federated learning: Challenges and future directions (2023) · cites this paper as [62]
6 more claims about this paper
Create a free account to see every claim other papers make about this one.
Citations
Cites
New directions in cryptography
Diffie, W., & Hellman, M. (1976) · IEEE Transactions on Information Theory · doi:10.1109/tit.1976.1055638 · Cited by 14,713
Communication-Efficient Learning of Deep Networks from Decentralized Data
McMahan, H. B., Moore, E., Ramage, D., Hampson, S., & Arcas, B. A. Y. (2016) · arXiv · doi:10.48550/arxiv.1602.05629 · Cited by 14,215 · cited 1×
“maintains a private database of her text messages securely on her own mobile device, and a shared global model is trained under the coordination of a central server based upon highly processed, minimally scoped, ephemeral updates from users [20], [10].”
BackgroundSelf-citationShamir, A. (1979) · Communications of the ACM · doi:10.1145/359168.359176 · Cited by 13,756 · cited 1×
“We rely on Shamir’s t -out-of- n Secret Sharing [38], which allows a user to split a secret s into n shares, such that any t shares can be used to reconstruct s , but any set of at most t − 1 shares gives no information about s .”
Method(2016) · Cited by 11,679
Deep Learning with Differential Privacy
Abadi, M., Chu, A., Goodfellow, I., McMahan, H. B., Mironov, I., Talwar, K., & Zhang, L. (2016) · Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security · doi:10.1145/2976749.2978318 · Cited by 6,578 · cited 3×
“This is particular advantageous in the local privacy setting [13], which offers provable guarantees for the protection of individual training examples [14], [15] even when the data aggregator is not assumed to be trusted [16], [17].”
BackgroundMethodSelf-citationThe Algorithmic Foundations of Differential Privacy
Dwork, C., & Roth, A. (2014) · Foundations and Trends® in Theoretical Computer Science · doi:10.1561/0400000042 · Cited by 5,557
Dwork, C. (2006) · Lecture Notes in Computer Science · doi:10.1007/11787006_1 · Cited by 5,291 · cited 1×
“In such cases, secure aggregation composes well with differential privacy [12].”
BackgroundMembership Inference Attacks Against Machine Learning Models
Shokri, R., Stronati, M., Song, C., & Shmatikov, V. (2017) · 2017 IEEE Symposium on Security and Privacy (SP) · doi:10.1109/sp.2017.41 · Cited by 4,754
The Algorithmic Foundations of Differential Privacy
Dwork, C., & Roth, A. (2013) · doi:10.1561/9781601988195 · Cited by 3,856
Federated Learning: Strategies for Improving Communication Efficiency
Konečný, J., McMahan, H. B., Yu, F. X., Richtárik, P., Suresh, A. T., & Bacon, D. (2016) · arXiv · doi:10.48550/arxiv.1610.05492 · Cited by 3,766
Self-citation
Research works printed but not linked: 3
- Stephen Advokat. 1987. Publication Of Bork's Video Rentals Raises Privacy Issue. Chicago Tribune (1987).
- Michael Barbaro, Tom Zeller, and Saul Hansell. 2006. A face is exposed for AOL searcher no. 4417749. New York Times 9, 2008 (2006).
- Elaine Shi, HTH Chan, Eleanor Rieffel, Richard Chow, and Dawn Song. 2011. Privacy-preserving aggregation of time-series data. In Annual Network & Distributed System Security Symposium (NDSS). Internet Society.
Cited by
Items 901–1,000 of 1,035
Federated Few-Shot Learning for Mobile NLP
Cai, D., Wang, S., Wu, Y., Lin, F. X., & Xu, M. (2023) · ACM Annual International Conference on Mobile Computing and Networking (MobiCom) · doi:10.1145/3570361.3613277 · Cited by 45
“Furthermore, integrating various privacy-preserving techniques, such as differential privacy [10] and secure aggregation [14], can further enhance the security of FL. FeS is parameter-efficient and thereby shall be easy to integrate with them.”
BackgroundFEVERLESS: Fast and Secure Vertical Federated Learning Based on XGBoost for Decentralized Labels
Wang, R., Ersoy, O., Zhu, H., Jin, Y., & Liang, K. (2022) · IEEE Transactions on Big Data · doi:10.1109/tbdata.2022.3227326 · Cited by 44
“(Wu et al. (2020)) introduced an approach for tree-based model training through a hybrid method composing homomorphic encryption and secure Multi-Party Computation (MPC) (Goldreich (1998); Bonawitz et al. (2017)).”
MethodEffectiveness of Distillation Attack and Countermeasure on Neural Network Watermarking
Yang, Z., Tang, T., Dang, H., Wu, Z., & Chang, E.-C. (2025) · IEEE Transactions on Dependable and Secure Computing · doi:10.1109/tdsc.2025.3609476 · Cited by 45
Stochastic, Distributed and Federated Optimization for Machine LearningPreprint
Konečný, J. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1707.01155 · Cited by 44
“In this regard, recent work proposed a cryptographic protocol for secure aggregation, ensuring that the server can only see an aggregate function of individual updates, such as an average, across many participating users [16].”
BackgroundUniversal Multi-Party Poisoning AttacksPreprint
Mahloujifar, S., Mahmoody, M., & Mohammed, A. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1809.03474 · Cited by 44
“In the distributed setting [MR17, MMR + 16, BIK + 17, KMY + 16], the training data T might be coming from various sources; e.g., it can be generated by m data providers P 1 , . . . , P m in an online way, while at the end a fixed algorithm, called the aggregator G , generates the hypothesis h based…”
BackgroundFedH2L: Federated Learning with Model and Statistical HeterogeneityPreprint
Li, Y., Zhou, W. G., Wang, H., Mi, H., & Hospedales, T. M. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2101.11296 · Cited by 44
“To enable collective benefit from AI while maintaining data privacy, Federated Learning (FL) [Bonawitz et al., 2017; McMahan et al., 2017; Konečný et al., 2016] algorithms aim to train a global model based on the efforts of distributed participants’ data and resources.”
BackgroundEfficient Sparse Secure Aggregation for Federated LearningPreprint
Béguier, C., Andreux, M., & Tramel, E. W. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2007.14861 · Cited by 56
“A standard protection technique is secure aggregation (Bonawitz et al., 2017; Dong et al., 2020).”
BackgroundMethodInfluentialAchieving Privacy-preserving Federated Learning with Irrelevant Updates over E-Health Applications
Chen, H., Li, H., Xu, G., Zhang, Y., & Luo, X. (2020) · ICC 2020 - 2020 IEEE International Conference on Communications (ICC) · doi:10.1109/icc40277.2020.9149385 · Cited by 43
“Bonawitz et al. [15] proposed a secure aggregation protocol by using secure multi-party computation.”
MethodResultRachakonda, S., Moorthy, S., Jain, A., Bukharev, A., Bucur, A., Manni, F., Quiterio, T. M., Joosten, L., & Mendez, N. I. (2023) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2022.3185418 · Cited by 43
“To enhance privacy when using FL in the healthcare domain, we employ secure multi-party computation (MPC) [26].”
MethodGao, Z., Wu, F., Gao, W., & Zhuang, X. (2022) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2022.3220750 · Cited by 43
“Since private information could be recovered from the model gradients or parameters during training, the previous works [16], [17] have proposed secured meth-ods for parameter aggregation.”
BackgroundRetracted: Communication-Efficient Personalized Federated Meta-Learning in Edge Networks
Yu, F., Lin, H., Wang, X., Garg, S., Kaddoum, G., Singh, S., & Hassan, M. M. (2023) · IEEE Transactions on Network and Service Management · doi:10.1109/tnsm.2023.3263831 · Cited by 43
“During FL training, there is a risk that model updates may unintentionally leak sensitive information to third-party or central servers [5].”
BackgroundRetractedTowards robust and privacy-preserving federated learning in edge computing
Zhou, H., Zheng, Y., & Jia, X. (2024) · Computer Networks · doi:10.1016/j.comnet.2024.110321 · Cited by 43
Rampone, G., Ivaniv, T., & Rampone, S. (2025) · Electronics · doi:10.3390/electronics14071430 · Cited by 43
“Differential privacy (DP) mitigates this by adding controlled noise to the updates, ensuring that the output of the aggregation process does not reveal too much about any single participant’s data [34].”
Robust and Secure Federated Learning With Verifiable Differential Privacy
Zhang, C., Weng, J., Weng, J., Zhong, Y., Liu, J.-N., & Deng, C. (2025) · IEEE Transactions on Dependable and Secure Computing · doi:10.1109/tdsc.2025.3574745 · Cited by 43
“Currently, numerous privacy-preserving mechanisms are employed to counter privacy leakage in FL, which can be broadly categorized into cryptographic primitives [12], [14], [21] and DP techniques [23], [25], [38].”
Multi-Agent Image Classification via Reinforcement Learning
Mousavi, H., Nazari, M., Takáč, M., & Motee, N. (2019) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros40897.2019.8968129 · Cited by 43
An accurate, scalable and verifiable protocol for federated differentially private averaging
Sabater, C., Bellet, A., & Ramon, J. (2022) · Machine Learning · doi:10.1007/s10994-022-06267-9 · Cited by 44
Federated Variational Autoencoder for Collaborative Filtering
Polato, M. (2021) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn52387.2021.9533358 · Cited by 43
“An off-the-shelf solution may be to employ a Secure Aggregation approach [36], [37].”
MethodEfficient Privacy-Preserving Federated Learning With Unreliable Users
Li, Y., Li, H., Xu, G., Huang, X., & Lu, R. (2021) · IEEE Internet of Things Journal · doi:10.1109/jiot.2021.3130115 · Cited by 43
“For addressing the privacy issues in FL, many state-of-the-art approaches have been proposed, mainly derived from three technologies, i.e., HE [12], [28], secure multiparty computation (MPC) [13], [29]–[31], and DP [32].”
BackgroundMethodInfluentialExplaining predictions and attacks in federated learning via random forests
Haffar, R., Sánchez, D., & Domingo‐Ferrer, J. (2022) · Applied Intelligence · doi:10.1007/s10489-022-03435-1 · Cited by 44
“Regarding privacy attacks by the model manager, these can be proactively prevented by the peers themselves by distorting their updates via differential privacy [39] or by securely aggregating updates of several peers before sending them to the server [40].”
BackgroundLaF: Lattice-Based and Communication-Efficient Federated Learning
Xu, P., Hu, M., Chen, T., Wang, W., & Jin, H. (2022) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2022.3186739 · Cited by 43
AGIC: Approximate Gradient Inversion Attack on Federated Learning
Jin, X., Hong, C., Huang, J., Chen, L. Y., & Decouchant, J. (2022) · IEEE Symposium on Reliable Distributed Systems (SRDS) · doi:10.1109/srds55811.2022.00012 · Cited by 44
“Federated learning (FL) [2], [5] is a collaborative learning paradigm that allows distributed data owners (called clients) to jointly learn machine learning models without centrally pooling their data.”
BackgroundZeph: Cryptographic Enforcement of End-to-End Data PrivacyPreprint
Burkhalter, L., Küchler, N., Viand, A., Shafagh, H., & Hithnawi, A. (2021) · arXiv (Cornell University) · doi:10.3929/ethz-b-000494008 · Cited by 43
Machine Unlearning of Federated ClustersPreprint
Pan, C., Sima, J., Prakash, S., Rana, V., & Milenković, O. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2210.16424 · Cited by 43
“, 2020), secure local model aggregation at the server is a critical consideration in FL, as it guarantees that the server cannot get specific information about client data based on their local models (Bonawitz et al., 2017; Bell et al., 2020; So et al., 2022; Chen et al., 2022).”
InfluentialZhao, D. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2410.13272 · Cited by 43
A systematic review of federated learning incentive mechanisms and associated security challenges
Ali, A., Ilahi, I., Qayyum, A., Mohammed, I., Al‐Fuqaha, A., & Qadir, J. (2023) · Computer Science Review · doi:10.1016/j.cosrev.2023.100593 · Cited by 43
MimiC: Combating Client Dropouts in Federated Learning by Mimicking Central Updates
Sun, Y., Mao, Y., & Zhang, J. (2023) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2023.3338021 · Cited by 42
“Notably, as the server only requires a sum ∑ i∈St v i t, secure aggregation techniques [43], [44] can be further implemented to preserve the privacy of local models.”
MethodShin, J., Yoon, H. S., Lee, S., Park, S., Liu, Y., Choi, J. D., & Lee, S.-J. (2023) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2023.emnlp-main.734 · Cited by 42
Towards the Robustness of Differentially Private Federated Learning
Qi, T., Wang, H., & Huang, Y. (2024) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v38i18.29967 · Cited by 42
ESVFL: Efficient and secure verifiable federated learning with privacy-preserving
Cai, J., Shen, W., & Qin, J. (2024) · Information Fusion · doi:10.1016/j.inffus.2024.102420 · Cited by 43
Over-the-Air Federated Learning in Digital Twins Empowered UAV Swarms
Jiang, B., Du, J., Jiang, C., Han, Z., Alhammadi, A., & Debbah, M. A. (2024) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2024.3455009 · Cited by 42
“Furthermore, the central server is unable to directly access the privacy-sensitive information of any individual device, as only the superimposed local model parameter signals can be obtained, aligning with the concept of secure aggregation [22].”
BackgroundFederated learning-based natural language processing: a systematic literature review
Khan, Y., Sánchez, D., & Domingo‐Ferrer, J. (2024) · Artificial Intelligence Review · doi:10.1007/s10462-024-10970-5 · Cited by 43
“Backdoor attacks in FL can be mitigated by secure aggregation algorithms using cryptographic techniques to ensure the privacy and security of the updates during the aggregation process and to detect any malicious updates that may contain backdoors (Bonawitz et al. 2017).”
Chen, L., Xiao, D., Xiao, X., & Zhang, Y. (2024) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2024.3486611 · Cited by 42
Chen, C., Zheng, F., Cui, J., Cao, Y., Liu, G., Wu, J. X., & Zhou, J. (2024) · International Journal of Machine Learning and Cybernetics · doi:10.1007/s13042-024-02106-6 · Cited by 42
“Moreover, secret sharing can be used with other privacy-enhancing technologies such as HE [31], PSI [16], FL [11], and random permutation [192], making it flexible to incorporate to a combination of privacy-preserving algorithms.”
Kang, S., Han, D., Lee, J., Im, D., Kim, S., Kim, S., Ryu, J., & Yoo, H. (2021) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2021.3066572 · Cited by 42
DistFL: Distribution-aware Federated Learning for Mobile Scenarios
Liu, B., Cai, Y., Zhang, Z., Li, Y., Wang, L., Li, D., Guo, Y., & Chen, X. (2021) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3494966 · Cited by 42
“Finally, although we have simulated the privacy-protection scenario by enforcing differential privacy (DP) to the local models, researchers may want to use some secure aggregation schemes [5] to further protect the model information.”
MethodThe Tradeoff Between Privacy and Accuracy in Anomaly Detection Using Federated XGBoostPreprint
Yang, M., Song, L., Xu, J., Li, C., & Tan, G. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1907.07157 · Cited by 42
“Take Paillier-based encryption schemes as an example, the cost of generating threshold decryption keys is very high [ Bonawitz et al. , 2017 ] .”
BackgroundEnhancing the Privacy of Federated Learning with SketchingPreprint
Liu, Z., Li, T., Smith, V., & Sekar, V. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1911.01812 · Cited by 43
“iscussed in detail in §2.1). Meanwhile, current efforts that attempt to boost the privacy of federated learning build upon traditional cryptographic techniques, such as secure multi-party computation [10] and differential privacy [8, 25]. As federated learning is commonly deployed on low-powered, mobile user devices,…”
BackgroundMethodInfluentialLayer-wise Characterization of Latent Information Leakage in Federated LearningPreprint
Mo, F., Borovykh, A., Malekzadeh, M., Haddadi, H., & Demetriou, S. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2010.08762 · Cited by 42
“There is a surge of interest in FL (Kairouz et al., 2019; Bonawitz et al., 2017; Hard et al., 2018), as it resolves the need for collecting private data in a centralized location which in many cases is not possible due to computational costs, privacy risks, and even legal considerations (Kairouz et al.”
InfluentialZero-Knowledge Proof-based Practical Federated Learning on BlockchainPreprint
Xing, Z., Zhang, Z., Li, M., Liu, J., Zhu, L., Russello, G., & Asghar, M. R. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2304.05590 · Cited by 42
“[28] proposed a one-time pad method to compute the sum X = Σi=1xi without revealing each party’s private factor xi to any other parties.”
MethodA Survey on Data MarketsPreprint
Zhang, J., Bi, Y., Cheng, M., Liu, J., Ren, K., Sun, Q., Wu, Y.-H., Cao, Y., Fernandez, R. C., Xu, H., et al. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2411.07267 · Cited by 42
Secure Partial Aggregation: Making Federated Learning More Robust for Industry 4.0 Applications
Gao, J., Zhang, B., Guo, X., Baker, T., Li, M., & Liu, Z. (2022) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2022.3145837 · Cited by 41
A federated learning scheme meets dynamic differential privacy
Guo, S., Wang, X., Long, S., Liu, H., Liu, H., & Sam, T. H. (2023) · CAAI Transactions on Intelligence Technology · doi:10.1049/cit2.12187 · Cited by 42
SAFARI: Sparsity-Enabled Federated Learning With Limited and Unreliable Communications
Mao, Y., Zhao, Z., Yang, M., Liang, L., Liu, Y., Ding, W., Lan, T., & Zhang, X.-P. S. (2023) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2023.3296624 · Cited by 41
Xie, C., Long, Y., Chen, P., Li, Q., Koyejo, S., & Li, B. (2023) · ACM SIGSAC Conference on Computer and Communications Security (CCS) · doi:10.1145/3576915.3623193 · Cited by 41
“As a result, approaches based on di � eren-tial privacy (DP) [22], homomorphic encryption [66], and secure multiparty computation [7, 13] have been proposed to protect the privacy of users in FL.”
BackgroundReguieg, H., Hanjri, M. E., Kamili, M. E., & Kobbane, A. (2023) · International Conference on Wireless Networks and Mobile Communications (WINCOM) · doi:10.1109/wincom59760.2023.10322899 · Cited by 41
“This algorithm is particularly effective in scenarios where data is unbalanced and non-IID (Independent and Identically Distributed), and it can significantly reduce the rounds of communication needed to train a deep network on decentralized data [8].”
BackgroundMethodInfluentialBPFL: Blockchain-based privacy-preserving federated learning against poisoning attack
Ren, Y., Hu, M., Yang, Z., Feng, G., & Zhang, X. (2024) · Information Sciences · doi:10.1016/j.ins.2024.120377 · Cited by 41
Survey of Federated Learning Models for Spatial-Temporal Mobility Applications
Belal, Y., Mokhtar, S. B., Haddadi, H., Wang, J., & Mashhadi, A. (2024) · ACM Transactions on Spatial Algorithms and Systems · doi:10.1145/3666089 · Cited by 41
“Other challenges within FL include the protection of user data and model parameters through methods such as differential privacy [74], Trusted Execution Environments [77], or encryption [12, 39].”
A Privacy-Preserving Federated Learning Framework With Lightweight and Fair in IoT
Chen, Y., Liu, L., Ping, Y., Atiquzzaman, M., Mumtaz, S., Zhang, Z., Guizani, M., & Tian, Z. (2024) · IEEE Transactions on Network and Service Management · doi:10.1109/tnsm.2024.3418786 · Cited by 41
“Bonawitz et al. [22] introduced a secure data aggregation scheme for machine learning utilizing secret sharing.”
BackgroundReis, M. J. C. S. (2025) · Applied Sciences · doi:10.3390/app15126452 · Cited by 43
An Emerging Decentralized Services Computing Paradigm for Big Data Governance: A Position Paper
Liu, X., Sun, X., & Huang, G. (2019) · IEEE Transactions on Services Computing · doi:10.1109/tsc.2019.2951558 · Cited by 41
“Differential privacy is considered to be helpful in preserving privacy for machine learning tasks, such as boosting, principal component analysis, linear and logistic regression, support vector machines, risk minimization, and continuous data processing [42].”
BackgroundZhang, J., & Simeone, O. (2020) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2020.2979762 · Cited by 41
“Lastly, a related issue would be to introduce data privacy requirements [48].”
BackgroundEdge Intelligence Assisted Gateway Defense in Cyber Security
Xu, S., Qian, Y., & Hu, R. Q. (2020) · IEEE Network · doi:10.1109/mnet.011.1900407 · Cited by 41
“E vAluAtIons on n Etwork A noMAly d EtEctIon applying a semi-supervised learning approach perform robust network anomaly detection. first describe the”
MethodAsad, M., Moustafa, A., & Aslam, M. (2021) · Applied Soft Computing · doi:10.1016/j.asoc.2021.107235 · Cited by 41
Locally Differentially Private Sparse Vector Aggregation
Zhou, M., Wang, T., Chan, T.-H. H., Fanti, G. C., & Shi, E. (2022) · IEEE Symposium on Security and Privacy (S&P) · doi:10.1109/sp46214.2022.9833635 · Cited by 41
Empirical analysis of federated learning in heterogeneous environments
Abdelmoniem, A. M., Ho, C.-Y., Papageorgiou, P., & Canini, M. (2022) · Workshop on Machine Learning and Systems (EuroMLSys) · doi:10.1145/3517207.3526969 · Cited by 41
“Finally, the server aggregates, with the global model, the model updates sent by the clients (possibly via secure aggregation [8]) and then checkpoints the new global model.”
MethodFederated Learning-Based Privacy-Preserving Data Aggregation Scheme for IIoT
Fan, H., Huang, C., & Liu, Y. (2022) · IEEE Access · doi:10.1109/access.2022.3226245 · Cited by 41
Privacy-Preserving Electricity Theft Detection Based on Blockchain
Zhao, Z., Liu, Y., Zeng, Z., Chen, Z., & Zhou, H. (2023) · IEEE Transactions on Smart Grid · doi:10.1109/tsg.2023.3246459 · Cited by 41
“[24] proposed a secure aggregation scheme”
PrivFairFL: Privacy-Preserving Group Fairness in Federated LearningPreprint
Pentyala, S., Neophytou, N., Nascimento, A. C. A., Cock, M. D., & Farnadi, G. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2205.11584 · Cited by 41
Carter, J. T., Feddema, J., Kothe, D. B., Neely, R., Pruet, J., Stevens, R. L., Balaprakash, P., Beckman, P., Foster, I. T., Iskra, K. A., et al. (2023) · OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) · doi:10.2172/1986455 · Cited by 41
DHSA: efficient doubly homomorphic secure aggregation for cross-silo federated learning
Liu, Z., Chen, S., Ye, J., Fan, J., Li, H., & Li, X. (2022) · The Journal of Supercomputing · doi:10.1007/s11227-022-04745-4 · Cited by 40
“Consequently, privacy protection techniques including secure multiparty computation (SMC) [8–12], homomorphic encryption [13–17], and differential privacy [18, 19] are proposed to implement secure aggregation schemes to address such indirect information leakage issues.”
InfluentialReview on application progress of federated learning model and security hazard protection
Yang, A., Ma, Z., Zhang, C., Han, Y., Hu, Z., Zhang, W., Huang, X., & Wu, Y. (2022) · Digital Communications and Networks · doi:10.1016/j.dcan.2022.11.006 · Cited by 40
Conceptualizing the Secure Machine Learning Operations (SecMLOps) Paradigm
Zhang, X., & Jaskolka, J. (2022) · IEEE International Conference on Software Quality, Reliability and Security (QRS) · doi:10.1109/qrs57517.2022.00023 · Cited by 40
“Based on the design solution by the solution architect (i.e., role R2), they discover and implement the most suitable defences and mitigations [30] to realize secure data ingestion and management to result in desired datasets.”
BackgroundBelt and Braces: When Federated Learning Meets Differential Privacy
Ren, X., Yang, S., Zhao, C., McCann, J. A., & Xu, Z. (2024) · Communications of the ACM · doi:10.1145/3650028 · Cited by 41
“Despite recent progress in theory [7, 9], building practical FL systems while addressing the above impacts simultaneously is still challenging.”
Privacy-Preserving Federated Unlearning With Certified Client Removal
Liu, Z., Ye, H., Jiang, Y., Shen, J., Guo, J., Tjuawinata, I., & Lam, K. (2025) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2025.3555868 · Cited by 40
“Potential countermeasures may involve techniques such as consistency checks [36] and verifiable computation [37].”
Federated Learning in Healthcare: From Research to Real-World Deployment
Bakas, S., Li, X., Shah, P., & Roth, H. R. (2026) · Annual Review of Biomedical Engineering · doi:10.1146/annurev-bioeng-080125-041414 · Cited by 40
“For example, adding too much differentially private noise to the gradients can lead to degraded model performance (86, 88, 89).”
Privacy-Preserving Machine Learning Training in IoT Aggregation Scenarios
Zhu, L., Tang, X., Shen, M., Gao, F., Zhang, J., & Du, X. (2021) · IEEE Internet of Things Journal · doi:10.1109/jiot.2021.3060764 · Cited by 40
An Improved Federated Learning Algorithm for Privacy Preserving in Cybertwin-Driven 6G System
Yang, M., Wang, X., Qian, H. L., Zhu, Y., Zhu, H., Guizani, M., & Chang, V. I. C. (2022) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2022.3149516 · Cited by 41
CE-Fed: Communication efficient multi-party computation enabled federated learning
Kanagavelu, R., Wei, Q., Li, Z., Zhang, H., Samsudin, J., Yang, Y., Goh, R. S. M., & Wang, S. (2022) · Array · doi:10.1016/j.array.2022.100207 · Cited by 41
Exploiting Defenses against GAN-Based Feature Inference Attacks in Federated Learning
Luo, X., & Zhang, X. (2025) · ACM Transactions on Knowledge Discovery from Data · doi:10.1145/3719350 · Cited by 40
“Numerous defense methods have been proposed to mitigate the risks incurred by inference attacks, including dropout [39, 45, 47], model stacking [45], anomaly detection [3], differential privacy [3, 12, 15, 23, 48], secure multiparty computation [4, 41, 52, 59, 66], and InstaHide [20].”
InfluentialFedSKETCH: Communication-Efficient and Private Federated Learning via SketchingPreprint
Haddadpour, F., Karimi, B., Li, P., & Li, X. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2008.04975 · Cited by 40
“formed by injecting an additional layer of random noise in order to respect differential-privacy property of the method [39] or using cryptography based approaches under secure multi-party computation [5] framework. Another promising recent approach with a potential to tackle all major issues in Federated Learning sett…”
MethodUnderstanding Unintended Memorization in Language Models Under Federated Learning
Thakkar, O. D., Ramaswamy, S., Mathews, R., & Beaufays, F. (2021) · Proceedings of the Third Workshop on Privacy in Natural Language Processing · doi:10.18653/v1/2021.privatenlp-1.1 · Cited by 39
“…Learning (FL) (McMahan et al., 2017a; McMahan and Ramage, 2017) has emerged as a popular approach for training neural networks on a large corpus of decentralized on-device data (e.g., (Koneˇcný et al., 2016; Konecný et al., 2016; Bonawitz et al., 2017; Hard et al., 2018; Bonawitz et al., 2019)).”
BackgroundFederated Many-Task Bayesian Optimization
Zhu, H., Wang, X., & Jin, Y. (2023) · IEEE Transactions on Evolutionary Computation · doi:10.1109/tevc.2023.3279775 · Cited by 40
“However, FDD-EA cannot protect the privacy of the newly sampled data, and therefore, a privacy-preserving federated data-driven multiobjective evolutionary algorithm was suggested in [14] by constructing the AF on a randomly chosen client integrated with secure aggregation [15].”
MethodSVCA: Secure and Verifiable Chained Aggregation for Privacy-Preserving Federated Learning
Xia, Y., Liu, Y., Dong, S., Li, M., & Guo, C. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3363712 · Cited by 39
“Additionally, in [12], [13], and [17], a similar pairwise mask SS scheme as in [21] is employed.”
MethodResultInfluentialA Privacy-Preserving Federated Learning Scheme Against Poisoning Attacks in Smart Grid
Li, X., Wen, M., He, S., Lu, R., & Wang, L. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3365142 · Cited by 39
“Bonawitz et al. [13] employed a combination of 2327-4662 c (cid:2) 2024 IEEE.”
BackgroundGahlan, N., & Sethia, D. (2024) · Multimedia Tools and Applications · doi:10.1007/s11042-024-19467-3 · Cited by 39
Liu, X., Wang, J., Xiong, X., & Sun, H. (2024) · Neurocomputing · doi:10.1016/j.neucom.2024.128653 · Cited by 39
PQSF: post-quantum secure privacy-preserving federated learning
Zhang, X., Deng, H., Wu, R., Ren, J., & Ren, Y. (2024) · Scientific Reports · doi:10.1038/s41598-024-74377-6 · Cited by 39
“Bonawitz et al. 24 proposed a privacy-protective secure aggregation scheme based on a participant-server framework in 2017.”
MethodLandscape of machine learning evolution: privacy-preserving federated learning frameworks and tools
Nguyen, G., Díaz, J. S.-P., Calatrava, A., Berberi, L., Lytvyn, O., Kozlov, V. Y., Tran, V., Moltó, G., & García, Á. L. (2024) · Artificial Intelligence Review · doi:10.1007/s10462-024-11036-2 · Cited by 39
Shi, S., Wang, N., Xiao, Y., Zhang, C., Shi, Y., Hou, Y. T., & Lou, W. (2025) · Network and Distributed System Security Symposium · doi:10.14722/ndss.2025.240644 · Cited by 39
MIASec: Enabling Data Indistinguishability Against Membership Inference Attacks in MLaaS
Wang, C., Liu, G., Huang, H., Feng, W., Peng, K., & Wang, L. (2019) · IEEE Transactions on Sustainable Computing · doi:10.1109/tsusc.2019.2930526 · Cited by 39
“Bonawitz et al. [31] propose a multi-party computation based protocol for secure aggregation over high-dimensional data for distributed machine learning.”
BackgroundSecure Federated Learning in 5G Mobile Networks
Isaksson, M., & Norrman, K. (2020) · Globecom · doi:10.1109/globecom42002.2020.9322479 · Cited by 39
“DEF compared to [10] in the fully malicious case.”
InfluentialStrategies for Enhancing Training and Privacy in Blockchain Enabled Federated Learning
Kumar, S., Dutta, S., Chatturvedi, S., & Bhatia, M. P. S. (2020) · IEEE International Conference on Multimedia Big Data · doi:10.1109/bigmm50055.2020.00058 · Cited by 40
“Homomorphic encryption has also been used in other federated learning experiments to enhance the security of the system [20].”
MethodFederatedReverse: A Detection and Defense Method Against Backdoor Attacks in Federated Learning
Zhao, C., Wen, Y., Li, S., Liu, F., & Meng, D. (2021) · ACM Workshop on Information Hiding and Multimedia Security (IH&MMSec) · doi:10.1145/3437880.3460403 · Cited by 39
“Based on considerations for party data privacy, detection methods cannot aggregate data or access model updates [8, 23].”
BackgroundMethodCORK: A privacy-preserving and lossless federated learning scheme for deep neural network
Zhao, J., Zhu, H., Wang, F., Lu, R., Li, H., Tu, J., & Shen, J. (2022) · Information Sciences · doi:10.1016/j.ins.2022.04.052 · Cited by 39
ESB-FL: Efficient and Secure Blockchain-Based Federated Learning With Fair Payment
Chen, B., Zeng, H., Xiang, T., Guo, S., Zhang, T., & Liu, Y. (2022) · IEEE Transactions on Big Data · doi:10.1109/tbdata.2022.3177170 · Cited by 39
TCT: Convexifying Federated Learning Using Bootstrapped Neural Tangent Kernels
Yu, Y., Wei, A., Karimireddy, S. P., Ma, Y., & Jordan, M. I. (2022) · neural information processing systems · doi:10.52202/068431-2239 · Cited by 39
Threats and Defenses in the Federated Learning Life Cycle: A Comprehensive Survey and Challenges
Li, Y., Guo, Z., Yang, N., Chen, H., Yuan, D., & Ding, W. (2025) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2025.3563537 · Cited by 38
“To achieve SMPC, the study [171] proposed secure aggregated FL (namely PPML) based on a Secret Sharing and Key Agreement Protocol.”
Multi-institutional PET/CT image segmentation using federated deep transformer learning
Shiri, I., Razeghi, B., Sadr, A. V., Amini, M., Salimi, Y., Ferdowsi, S., Boor, P., Gündüz, D., Voloshynovskiy, S., & Zaidi, H. (2023) · Computer Methods and Programs in Biomedicine · doi:10.1016/j.cmpb.2023.107706 · Cited by 38
Machine learning cryptography methods for IoT in healthcare
Chinbat, T., Madanian, S., Airehrour, D., & Hassandoust, F. (2024) · BMC Medical Informatics and Decision Making · doi:10.1186/s12911-024-02548-6 · Cited by 38
“For instance, Federated Learning (FL), an ML subset, enables the training of a global ML model using data from multiple IoT devices without the need to share the data [26].”
VPFL: Enabling verifiability and privacy in federated learning with zero-knowledge proofs
Ma, J., Liu, H., Zhang, M., & Liu, Z. (2024) · Knowledge-Based Systems · doi:10.1016/j.knosys.2024.112115 · Cited by 38
A survey on privacy-preserving federated learning against poisoning attacks
Feng, X., & Cheng, W. (2024) · Cluster Computing · doi:10.1007/s10586-024-04629-7 · Cited by 38
“Previous studies [5–9] have revealed that attackers can infer sensitive information, such as training data samples, original model parameters, or functions, by analyzing gradients, even if they possess only a portion of the gradient.”
Grataloup, A., & Kurpicz-Briki, M. (2024) · Frontiers in Digital Health · doi:10.3389/fdgth.2024.1495999 · Cited by 38
Secure and Verifiable Inference in Deep Neural Networks
Xu, G., Li, H., Ren, H., Sun, J., Xu, S., Ning, J., Yang, H., Yang, K., & Deng, R. H. (2020) · Annual Computer Security Applications Conference · doi:10.1145/3427228.3427232 · Cited by 38
“Existing privacy-preserving methods for neural networks are mostly evolved from three underlying techniques: Secure Multi-Party Computation (SMC) [4, 35], Differential Privacy [45, 62] and Homomorphic Encryption (HE) [2, 55].”
BackgroundMethodInfluentialConcentrated Differentially Private Federated Learning With Performance Analysis
Hu, R., Guo, Y., & Gong, Y. (2021) · IEEE Open Journal of the Computer Society · doi:10.1109/ojcs.2021.3099108 · Cited by 40
“In the following, we provide a customized secure aggregation protocol similar as [25], which is efficient in terms of the amortized computation and communication overhead across all communication rounds.”
MethodResultInfluentialPreserving data privacy in federated learning through large gradient pruning
Zhang, Z., Zhu, T., Ren, W., Xiong, P., & Choo, K. R. (2022) · Computers & Security · doi:10.1016/j.cose.2022.103039 · Cited by 38
Federated learning‐based trajectory prediction model with privacy preserving for intelligent vehicle
Han, M., Xu, K., Ma, S., Li, A., & Jiang, H. (2022) · International Journal of Intelligent Systems · doi:10.1002/int.22987 · Cited by 37
“13,32 Although this can be achieved by using homomorphic encryption, but the encryption and decryption keys are still under the threat of being intercepted during the key distribution process, which will lead to the leakage of the whole model data.”
BackgroundHorizontal Federated Recommender System: A Survey
Wang, L., Zhou, H., Bao, Y., Yan, X., Shen, G., & Kong, X. (2024) · ACM Computing Surveys · doi:10.1145/3656165 · Cited by 37
“SMC consumes more data exchanges among participants when multiple privacy protocols cooperate with each other [ 11 , 14 ], and DP is not so resilient when confronted with existing malicious attacks [ 139 , 201 ].”
BackgroundMethodCrowdGuard: Federated Backdoor Detection in Federated Learning
Rieger, P., Krauß, T., Miettinen, M., Dmitrienko, A., & Sadeghi, A.-R. (2024) · Proceedings 2024 Network and Distributed System Security Symposium · doi:10.14722/ndss.2024.23233 · Cited by 36
“Secure Aggregation Techniques: Various approaches have been proposed to prevent honest-but-curious [26] or fully malicious servers [11], [60], [35] from accessing the local model updates.”
BackgroundPapadopoulos, C., Kollias, K.-F., & Fragulis, G. F. (2024) · Future Internet · doi:10.3390/fi16110415 · Cited by 35
“Mechanisms such as secure aggregation, which uses cryptographic protocols to mix model updates from different nodes [65], help to maintain privacy during the aggregation process.”
Federated learning vulnerabilities, threats and defenses: A systematic review and future directions
Almutairi, S., & Barnawi, A. (2023) · Internet of Things · doi:10.1016/j.iot.2023.100947 · Cited by 35
Metadata
- Authors
- Bonawitz, K., Ivanov, V., Kreuter, B., Marcedone, A., McMahan, H. B., Patel, S., Ramage, D., Segal, A., & Seth, K.
- Affiliations
- Google (United States) (US)
- Cornell Tech (US)
- Venue
- Proceedings of the 2017 ACM SIGSAC Conference on Computer and Communications Security
- Publisher
- ACM
- Published
- 2017-10-30
- Volume
- 2017
- Pages
- 1175-1191
- ISSN
- 1543-7221 (ISSN-L)
- OpenAlex
- W2767079719
- Cited by
- 3,956
- References
- 55
- Citations by year
- 2026: 701 · 2025: 712 · 2024: 643 · 2023: 564 · 2022: 400 · 2021: 436 · 2020: 299 · 2019: 152 · 2018: 42 · 2017: 7
- FWCI
- 62.37
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- ACM SIGSAC Conference on Computer and Communications Security (CCS) (publishedVersion)
- IACR Cryptology ePrint Archive (submittedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Networks and Communications › Cryptography and Data Security
- Physical Sciences › Computer Science › Artificial Intelligence › Privacy-Preserving Technologies in Data
- Physical Sciences › Computer Science › Information Systems › Cloud Data Security Solutions
- Keywords
- secure aggregation, privacy-preserving machine learning, federated learning, high-dimensional data, communication efficiency, user dropout, honest-but-curious adversary, active adversaries, secure multi-party computation
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.