Mitigating communications threats in decentralized federated learning through moving target defense

Wireless Networks
Enrique Tomás Martínez Beltrán1, Pedro Miguel Sánchez Sánchez1, Sergio López Bernal1, Gérôme Bovet2, Manuel Gil Pérez1, Gregorio Martínez Pérez1, Alberto Huertas Celdrán3
1Department of Information and Communications Engineering, University of Murcia, 30100 Murcia, Spain
2Cyber-Defence Campus, Armasuisse Science and Technology, 3602, Thun, Switzerland
3Communication Systems Group, Department of Informatics (IFI), University of Zurich, 8050, Zürich, Switzerland

Tóm tắt

AbstractThe rise of Decentralized Federated Learning (DFL) has enabled the training of machine learning models across federated participants, fostering decentralized model aggregation and reducing dependence on a server. However, this approach introduces unique communication security challenges that have yet to be thoroughly addressed in the literature. These challenges primarily originate from the decentralized nature of the aggregation process, the varied roles and responsibilities of the participants, and the absence of a central authority to oversee and mitigate threats. Addressing these challenges, this paper first delineates a comprehensive threat model focused on DFL communications. In response to these identified risks, this work introduces a security module to counter communication-based attacks for DFL platforms. The module combines security techniques such as symmetric and asymmetric encryption with Moving Target Defense (MTD) techniques, including random neighbor selection and IP/port switching. The security module is implemented in a DFL platform, Fedstellar, allowing the deployment and monitoring of the federation. A DFL scenario with physical and virtual deployments have been executed, encompassing three security configurations: (i) a baseline without security, (ii) an encrypted configuration, and (iii) a configuration integrating both encryption and MTD techniques. The effectiveness of the security module is validated through experiments with the MNIST dataset and eclipse attacks.The results showed an average F1 score of 95%, with the most secure configuration resulting in CPU usage peaking at 68% (± 9%) in virtual deployments and network traffic reaching 480.8 MB (± 18 MB), effectively mitigating risks associated with eavesdropping or eclipse attacks.

Từ khóa


Tài liệu tham khảo

Reinsel, D., Gantz, J., & Rydnin, J.: The digitization of the world from edge to core (2018). https://www.seagate.com/files/www-content/our-story/trends/files/idc-seagate-dataage-whitepaper.pdf

Martínez Beltrán, E. T., Quiles Pérez, M., Sánchez Sánchez, P. M., López Bernal, S., Bovet, G., Gil Pérez, M., Martínez Pérez, G., & Huertas Celdrán, A. (2023). Decentralized federated learning: Fundamentals, state of the art, frameworks, trends, and challenges. IEEE Communications Surveys & Tutorials, 25(4), 2983–3013. https://doi.org/10.1109/COMST.2023.3315746

Shi, Y., Liu, Y., Sun, Y., Lin, Z., Shen, L., Wang, X., & Tao, D.: Towards more suitable personalization in federated learning via decentralized partial model training. arXiv preprint arXiv:2305.15157 (2023)

Salama, A., Stergioulis, A., Hayajneh, A. M., Zaidi, S. A. R., McLernon, D., & Robertson, I. (2023). Decentralized federated learning over slotted aloha wireless mesh networking. IEEE Access, 11, 18326–18342. https://doi.org/10.1109/ACCESS.2023.3246924

Xiao, Y., Ye, Y., Huang, S., Hao, L., Ma, Z., Xiao, M., Mumtaz, S., & Dobre, O. A. (2021). Fully decentralized federated learning-based on-board mission for UAV swarm system. IEEE Communications Letters, 25(10), 3296–3300. https://doi.org/10.1109/LCOMM.2021.3095362

Perales Gómez, Á.L., Martínez Beltrán, E. T., Sánchez Sánchez, P. M., & Huertas Celdrán, A.: TemporalFED: detecting cyberattacks in industrial time-series data using decentralized federated learning. arXiv preprint arXiv:2308.03554 (2023)

Mothukuri, V., Parizi, R. M., Pouriyeh, S., Huang, Y., Dehghantanha, A., & Srivastava, G. (2021). A survey on security and privacy of federated learning. Future Generation Computer Systems, 115, 619–640. https://doi.org/10.1016/j.future.2020.10.007

Etxezarreta, X., Garitano, I., Iturbe, M., & Zurutuza, U. (2023). Low delay network attributes randomization to proactively mitigate reconnaissance attacks in industrial control systems. Wireless Networks. https://doi.org/10.1007/s11276-022-03212-5

Gholami, A., Torkzaban, N., & Baras, J. S.: Trusted decentralized federated learning. In IEEE 19th annual consumer communications and networking conference (pp. 1–6) (2022). https://doi.org/10.1109/CCNC49033.2022.9700624

Mothukuri, V., Parizi, R. M., Pouriyeh, S., Dehghantanha, A., & Choo, K.-K.R. (2022). FabricFL: Blockchain-in-the-loop federated learning for trusted decentralized systems. IEEE Systems Journal, 16(3), 3711–3722. https://doi.org/10.1109/JSYST.2021.3124513

Li, Y., Wang, X., Sun, R., Xie, X., Ying, S., & Ren, S. (2023). Trustiness-based hierarchical decentralized federated learning. Knowledge-Based Systems. https://doi.org/10.1016/j.knosys.2023.110763

Chen, Y., Liang, L., & Gao, W. (2023). Non trust detection of decentralized federated learning based on historical gradient. Engineering Applications of Artificial Intelligence, 120, 105888. https://doi.org/10.1016/j.engappai.2023.105888

Wang, P., Sun, W., Zhang, H., Ma, W., & Zhang, Y. (2023). Distributed and secure federated learning for wireless computing power networks. IEEE Transactions on Vehicular Technology. https://doi.org/10.1109/TVT.2023.3247859

Singh, S. K., Yang, L. T., & Park, J. H. (2023). Fusionfedblock: Fusion of blockchain and federated learning to preserve privacy in industry 5.0. Information Fusion, 90, 233–240. https://doi.org/10.1016/j.inffus.2022.09.027

Qu, Y., Xu, C., Gao, L., Xiang, Y., & Yu, S. (2022). FL-SEC: Privacy-preserving decentralized federated learning using SignSGD for the Internet of Artificially Intelligent Things. IEEE Internet of Things Magazine, 5(1), 85–90. https://doi.org/10.1109/IOTM.001.2100173

Wang, L., Zhao, X., Lu, Z., Wang, L., & Zhang, S. (2023). Enhancing privacy preservation and trustworthiness for decentralized federated learning. Information Sciences, 628, 449–468. https://doi.org/10.1016/j.ins.2023.01.130

Arapakis, I., Papadopoulos, P., Katevas, K., & Perino, D.: P4l: Privacy preserving peer-to-peer learning for infrastructureless setups. arXiv preprint arXiv:2302.13438 (2023)

Ridhawi, I. A., Otoum, S., & Aloqaily, M.: Decentralized zero-trust framework for digital twin-based 6g. arXiv preprint arXiv:2302.03107 (2023)

Martínez Beltrán, E. T., Perales Gómez, Á. L., Feng, C., Sánchez Sánchez, P. M., López Bernal, S., Bovet, G., Gil Pérez, M., Martínez Pérez, G., & Huertas Celdrán, A. (2024). Fedstellar: A platform for decentralized federated learning. Expert Systems with Applications, 242, 122861. https://doi.org/10.1016/j.eswa.2023.122861

Deng, L. (2012). The MNIST database of handwritten digit images for machine learning research. IEEE Signal Processing Magazine, 29(6), 141–142. https://doi.org/10.1109/MSP.2012.2211477

Alangot, B., Reijsbergen, D., Venugopalan, S., Szalachowski, P., & Yeo, K. S. (2021). Decentralized and lightweight approach to detect eclipse attacks on proof of work blockchains. IEEE Transactions on Network and Service Management, 18(2), 1659–1672. https://doi.org/10.1109/TNSM.2021.3069502

Niu, B., Chen, Y., Wang, Z., Li, F., Wang, B., & Li, H. (2022). Eclipse: Preserving differential location privacy against long-term observation attacks. IEEE Transactions on Mobile Computing, 21(1), 125–138. https://doi.org/10.1109/TMC.2020.3000730

Thông tin
Thông tin xuất bản

Wireless Networks

Thông tin tác giả