BFL-Guard: A Blockchain-Enabled Federated Learning Framework with Zero-Knowledge Gradient Verification and Tokenized Incentives
Article Sidebar
Main Article Content
Federated Learning (FL) enables collaborative model training across decentralized participants without sharing raw data. However, existing FL systems remain vulnerable to Byzantine attacks and suffer from a lack of accountability, verifiability, and economic incentives for honest participation. We present BFL-Guard, a novel blockchain-orchestrated federated learning framework integrating: (i) zk-SNARK-based zero-knowledge gradient proofs, (ii) an on-chain Byzantine-tolerant aggregation smart contract, and (iii) a tokenized incentive protocol (FedToken). BFL-Guard stores model checkpoints as IPFS hashes anchored on Ethereum, ensuring tamper-evident auditability. Experiments on CIFAR-10 and Shakespeare benchmarks demonstrate 95.2% and 87.6% accuracy in IID and Non-IID settings, surpassing all baselines while converging 12.4% faster even under 30% Byzantine injection.
Downloads
References
McMahan, H.B., Moore, E., Ramage, D., Hampson, S., & Agüera y Arcas, B. (2017). Communication-efficient learning of deep networks from decentralized data. AISTATS 2017.
Blanchard, P., El Mhamdi, E.M., Guerraoui, R., & Stainer, J. (2017). Machine learning with adversaries: Byzantine tolerant gradient descent. NeurIPS 2017.
Yin, D., Chen, Y., Kannan, R., & Bartlett, P. (2018). Byzantine-robust distributed learning: Towards optimal statistical rates. ICML 2018.
Bao, X., Su, C., Xiong, Y., Huang, W., & Hu, Y. (2019). FLChain: A blockchain for auditable federated learning with trust and incentive. IEEE BigDataService 2019.
Kim, H., Park, J., Bennis, M., & Kim, S.L. (2020). Blockchained on-device federated learning. IEEE Communications Letters, 24(6), 1279–1283.
Nguyen, D.C., Ding, M., Pathirana, P.N., et al. (2021). Federated learning for internet of things: A comprehensive survey. IEEE Communications Surveys & Tutorials, 23(3), 1622–1658.
Garg, S., Gentry, C., Halevi, S., & Raykova, M. (2023). Verifiable neural network inference using zk-SNARKs. IEEE S&P 2023.
Sun, G., Cong, Y., Dong, J., et al. (2022). Data poisoning attacks on federated machine learning. IEEE Internet of Things Journal, 9(13), 11365–11375.
Fang, M., Cao, X., Jia, J., & Gong, N.Z. (2020). Local model poisoning attacks to Byzantine-robust federated learning. USENIX Security 2020.
Groth, J. (2016). On the size of pairing-based non-interactive arguments. EUROCRYPT 2016.
Ethereum Improvement Proposal 4844: Shard Blob Transactions. https://eips.ethereum.org/EIPS/eip-4844 (2023).
Wood, G. (2014). Ethereum: A secure decentralised generalised transaction ledger. Ethereum Project Yellow Paper.
Osborne, M.J., & Rubinstein, A. (1994). A Course in Game Theory. MIT Press.
Zhan, Y., Zhang, J., Hong, Z., Wu, L., Li, P., & Guo, S. (2022). A Survey of Incentive Mechanism Design for Federated Learning. IEEE Transactions on Emerging Topics in Computing, 10(2), 1035–1044.
This work is licensed under a Creative Commons Attribution 4.0 International License.
All articles published in our journal are licensed under CC-BY 4.0, which permits authors to retain copyright of their work. This license allows for unrestricted use, sharing, and reproduction of the articles, provided that proper credit is given to the original authors and the source.