Abstract
Machine learning techniques are used in numerous applications to identify complex patterns and relationships in data. However, data of a single actor is often insufficient, as large amounts of data are required to train powerful machine learning models. One approach to tackle this problem is the federated training of models by multiple cooperating entities. However, the cooperation raises security and privacy concerns, especially if competitors are involved that do not want to share business critical training data with each other. We present a secure, privacy preserving, decentralized P2P Federated Learning framework to address these issues. The framework eliminates the need to establish a central trusted server for model training, which often represents a communication bottleneck or single point of failure. Various steps of data preprocessing, as well as the aggregation of individual models are carried out by means of Secure Multi-Party Computation to protect the data of training participants. We describe our experiments to demonstrate the basic feasibility of the working prototype and highlight open technical and methodological issues that we aim to address in the future.
Funded by the German Federal Ministry of Education and Research. Project name: KIWI, RefNr: 16KIS1142K.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Abadi, M., et al.: Deep learning with differential privacy. In: Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security, pp. 308–318 (2016)
Aïvodji, U., Gambs, S., Ther, T.: GAMIN: an adversarial approach to black-box model inversion. CoRR abs/1909.11835 (2019)
Bellet, A., Guerraoui, R., Taziki, M., Tommasi, M.: Personalized and private peer-to-peer machine learning. In: AISTATS, vol. 84, pp. 473–481. PMLR (2018)
Bonawitz, K., et al., V.I.: Practical secure aggregation for privacy-preserving machine learning. In: Thuraisingham, B.M., et al., D.E. (eds.) Proceedings of the 2017 ACM SIGSAC Conference on Computer and Communications Security, CCS, pp. 1175–1191. ACM (2017)
Chou, L., Liu, Z., Wang, Z., Shrivastava, A.: Efficient and less centralized federated learning. In: Oliver, N., Pérez-Cruz, F., Kramer, S., Read, J., Lozano, J.A. (eds.) ECML PKDD 2021. LNCS, vol. 12975, pp. 772–787. Springer, Cham (2021). https://doi.org/10.1007/978-3-030-86486-6_47
Cong, K., et al.: Labeled psi from homomorphic encryption with reduced computation and communication. Cryptology ePrint Archive, Report 2021/1116 (2021). https://ia.cr/2021/1116
Hard, A., et al.: Federated learning for mobile keyboard prediction. CoRR abs/1811.03604 (2018)
He, Z., Zhang, T., Lee, R.B.: Model inversion attacks against collaborative inference. In: Proceedings of the 35th Annual Computer Security Applications Conference, pp. 148–162. ACSAC 2019, Association for Computing Machinery, New York, NY, USA (2019). https://doi.org/10.1145/3359789.3359824
Kairouz, P., McMahan, H.B., Avent, B., Bellet, A., Bennis, M., et al.: Advances and open problems in federated learning. CoRR abs/1912.04977 (2019). http://arxiv.org/abs/1912.04977
Konečný, J., McMahan, H.B., Yu, F.X., Richtarik, P., Suresh, A.T., Bacon, D.: Federated learning: strategies for improving communication efficiency. In: NIPS Workshop on Private Multi-Party Machine Learning (2016)
Melis, L., Song, C., Cristofaro, E.D., Shmatikov, V.: Inference attacks against collaborative learning. CoRR abs/1805.04049 (2018). http://arxiv.org/abs/1805.04049
Patarasuk, P., Yuan, X.: Bandwidth optimal all-reduce algorithms for clusters of workstations. J. Parallel Distrib. Comput. 69(2), 117–124 (2009). https://doi.org/10.1016/j.jpdc.2008.09.002
Paverd, A.J., Martin, A.C.: Modelling and automatically analysing privacy properties for honest-but-curious adversaries (2014)
Reddi, S.J., et al.: Adaptive federated optimization. In: International Conference on Learning Representations (2021)
Rigaki, M., Garcia, S.: A survey of privacy attacks in machine learning (2021)
Roy, A.G., Siddiqui, S., Pölsterl, S., Navab, N., Wachinger, C.: BrainTorrent: a peer-to-peer environment for decentralized federated learning. CoRR abs/1905.06731 (2019)
Salem, A., Bhattacharya, A., Backes, M., Fritz, M., Zhang, Y.: Updates-leak: data set inference and reconstruction attacks in online learning. In: 29th USENIX Security Symposium (USENIX Security 20), pp. 1291–1308. USENIX Association (2020), https://www.usenix.org/conference/usenixsecurity20/presentation/salem
Shayan, M., Fung, C., Yoon, C.J.M., Beschastnikh, I.: Biscotti: a ledger for private and secure peer-to-peer machine learning. CoRR abs/1811.09904 (2018). http://arxiv.org/abs/1811.09904
Wink, T., Nochta, Z.: An approach for peer-to-peer federated learning. In: Proceedings of 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks Workshops (DSN-W) (2021)
Wittkopp, T., Acker, A.: Decentralized federated learning preserves model and data privacy. In: Hacid, H., et al. (eds.) ICSOC 2020. LNCS, vol. 12632, pp. 176–187. Springer, Cham (2021). https://doi.org/10.1007/978-3-030-76352-7_20
Zhao, H., Canny, J.F.: Sparse allreduce: efficient scalable communication for power-law data. CoRR abs/1312.3020 (2013). http://arxiv.org/abs/1312.3020
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Piotrowski, T., Nochta, Z. (2022). Towards a Secure Peer-to-Peer Federated Learning Framework. In: Zirpins, C., et al. Advances in Service-Oriented and Cloud Computing. ESOCC 2022. Communications in Computer and Information Science, vol 1617. Springer, Cham. https://doi.org/10.1007/978-3-031-23298-5_2
Download citation
DOI: https://doi.org/10.1007/978-3-031-23298-5_2
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-23297-8
Online ISBN: 978-3-031-23298-5
eBook Packages: Computer ScienceComputer Science (R0)