IEEE Transactions on Information Forensics and Security · 2022 · 308 citations · 22 references
ComputingComputer ScienceCybersecurityComputer SecurityCryptographyMachine LearningDifferential PrivacyFederated LearningFederated Learning PrivacyHomomorphic EncryptionPrivacy-preserving ProtocolsBenchmark DatasetsByzantine-resilient Federated LearningCosine SimilarityGradient Leakage AttacksPrivacy-preserving LearningPrivacy-preserving Machine LearningPrivacy-preserving Model Training
Privacy-Preserving Federated Learning (PPFL) is an emerging secure distributed learning paradigm that aggregates user-trained local gradients into a federated model through a cryptographic protocol. Unfortunately, PPFL is vulnerable to model poisoning attacks launched by a Byzantine adversary, who crafts malicious local gradients to harm the accuracy of the federated model. To resist model poisoning attacks, existing defense strategies focus on identifying suspicious local gradients over plaintexts. However, the Byzantine adversary submits encrypted poisonous gradients to circumvent existing defense strategies in PPFL, resulting in encrypted model poisoning. To address the issue, in this paper we design a privacy-preserving defense strategy using two-trapdoor homomorphic encryption (referred to as ShieldFL), which can resist encrypted model poisoning without compromising privacy in PPFL. Specially, we first present the secure cosine similarity method aiming to measure the distance between two encrypted gradients. Then, we propose the Byzantine-tolerance aggregation using cosine similarity, which can achieve robustness for both Independently Identically Distribution (IID) and non-IID data. Extensive evaluations on three benchmark datasets ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">i.e.,</i> MNIST, KDDCup99, and Amazon) show that ShieldFL outperforms existing defense strategies. Especially, ShieldFL can achieve 30%-80% accuracy improvement to defend two state-of-the-art model poisoning attacks in both non-IID and IID settings.
22
Membership Inference Attacks Against Machine Learning Models
Reza Shokri, Marco Stronati, Congzheng Song et al. · 2017 · 4.2K citations
Practical Secure Aggregation for Privacy-Preserving Machine Learning
Keith Bonawitz, Vladimir Ivanov, Ben Kreuter et al. · 2017 · 3.4K citations
Privacy-Preserving Deep Learning
Reza Shokri, Vitaly Shmatikov · 2015 · 2.3K citations
Federated Learning with Non-IID Data
Yue Zhao, Meng Li, Liangzhen Lai et al. · arXiv (Cornell University) · 2018 · 1.9K citations · Full text
SecureML: A System for Scalable Privacy-Preserving Machine Learning
Payman Mohassel, Yupeng Zhang · 2017 · 1.8K citations