Abstract
With the arrival of the Web3 era, data has seen an explosive growth. The use of differential privacy mechanisms in federated learning has been proposed to protect user privacy and avoid security threats from data sharing. The core idea of this approach is that multiple clients train their local data, add noise to their client parameters, and then transmit them to a central server for parameter aggregation. However, there are still defects that need to be addressed. First, it is difficult to resist attacks from malicious clients, which means that user privacy is not fully protected. Second, it is challenging to add an appropriate amount of noise to achieve high model accuracy. Therefore, this paper proposes a bidirectional adaptive noise addition federated learning scheme, which adds adaptive noise satisfying the differential privacy mechanism to both the central server and clients to improve model accuracy. Considering the heterogeneity of client hardware, this paper samples gradients and samples separately to reduce communication costs and uses RMSprop to accelerate model training on both clients and central servers. Experimental results show that the proposed scheme enhances user privacy protection while maintaining high efficiency.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Artificial Intelligence Project Team of Webank. Federated Learning White Paper V2.0. https://aisp-1251170195.cos.ap-hongkong.myqcloud.com/wp-content/uploads/pdf. Accessed 13 Feb 2023
Mcmahan, H.B.F., Moore, E,.S., Ramage D, T.: Communication-efficient learning of deep networks from decentralized data. In: Artificial Intelligence and Statistics. PMLR, pp. 1273–1282 (2017)
Tu, X.F., Zhu K.S., Luong N.C.T.: Incentive mechanisms for federated learning: from economic and game theoretic perspective. CoRR abs, 2111.11850 (2021)
Chen, J.F., Sun, C.S., Zhou, X.T.: Local privacy protection for power data prediction model based on federated learning and homomorphic encryption. Inf. Secur. Res. 9(03), 228–234(2023)
Su, Y.F., Liu, W.S.: Secure protection method for federated learning model based on secure shuffling and differential privacy. Inf. Secur. Res. 8(03), 270–276 (2022)
Li, T.F., Sahu, A.K.S., Talwalkar, A.T.: Federated learning: challenges, methods, and future directions. IEEE Sig. Process. Mag. 37, 50–60 (2020)
Liu, W.F., Chen, L.S., Chen Y.T.: Accelerating federated learning via momentum gradient descent. IEEE Trans. Parallel Distrib. Syst. 31(8), 1754–66 (2020)
Liu, X.F., Li, Y.S., Wang, Q,T.: Sparse personalized federated learning via maximizing correlation. ar**v preprint, ar**v.2107.05330 (2021)
Melis, L.F., Song C.S., Cristofaro, E.D.T.: Inference attacks against collaborative learning. CoRR abs, 1805.04049 (2018)
Abadi, M.F., Chu, A.S.: Deep learning with differential privacy. In: Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security, pp. 308–318 (2016)
Wei, K.F., Li, J.S., Ding, M.T.: User-level privacy-preserving federated learning: analysis and performance optimization. IEEE Trans. Mob. Comput. 21(9), 3388–3401 (2022)
Wu, X.F., Zhang, Y.S., Shi, M.T.: An adaptive federated learning scheme with differential privacy preserving. Future Gener. Comput. Syst. 127, 362–72 (2022)
Roth, A.F., Dwork, C.S.: The algorithmic foundations of differential privacy. 1st edn. Found. Trends Theor. Comput. Sci. 9(3–4), 211–407 (2013)
Li, N.F., Lyu, M.S., Su, D.T.: Differential privacy: from theory to practice. Synth. Lect. Inf. Secur. Priv. Trust 8(4), 1–138 (2016)
Geyer, R.C.F., Klein, T.S., Nabi, M.T.: Differentially private federated learning: a client level perspective. ar**v preprint, ar**v.1712.07557 (2017)
**ang, L.F., Yang, J.S., Li, B.T.: Differentially-private deep learning from an optimization perspective. In: Proceedings of the IEEE Conference on Computer Communications, pp. 559–567 (2019)
Acknowledgments
This research is funded by the 2022 Central University of Finance and Economics Education and Teaching Reform Fund (No. 2022ZXJG35), Emerging Interdisciplinary Project of CUFE, the National Natural Science Foundation of China (No. 61906220) and Ministry of Education of Humanities and Social Science project (No. 19YJCZH178).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2024 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Li, Y., Xu, J., Zhu, J., Wang, X. (2024). An Optimized Scheme of Federated Learning Based on Differential Privacy. In: Chen, J., Wen, B., Chen, T. (eds) Blockchain and Trustworthy Systems. BlockSys 2023. Communications in Computer and Information Science, vol 1896. Springer, Singapore. https://doi.org/10.1007/978-981-99-8101-4_20
Download citation
DOI: https://doi.org/10.1007/978-981-99-8101-4_20
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-99-8100-7
Online ISBN: 978-981-99-8101-4
eBook Packages: Computer ScienceComputer Science (R0)