FedSBS: Seleção de Participantes Baseado em Pontuação para Aprendizado Federado no Cenário de Detecção de Intrusão

• Helio N. Cunha Neto UFF
• Natalia C. Fernandes UFF
• Diogo M. F. Mattos UFF

DOI: https://doi.org/10.5753/sbseg.2023.233639

Abstract

Intrusion Detection Systems based on Federated Learning pose challenges for cybersecurity, including managing imbalanced data and interference from malicious participants. Federated Learning is a collaborative approach to machine learning that allows participants to train models with their local data while preserving privacy. A global model aggregates local models. However, malicious participants can compromise the global model with random or biased data. This article proposes the FedSBS method for participant selection. FedSBS aims to score each participant’s contribution and then proceed with the selection process. The method seeks to minimize the risks posed by malicious participants while optimizing the performance of the global model. The proposal demonstrates superior performance compared to other participant selection methods, achieving 80% F1-score, 90% accuracy, and 69% precision in the test set with malicious participants. FedSBS maintains the performance of the global model even with up to 60% malicious participants.

References

Agrawal, S., Sarkar, S., Aouedi, O., Yenduri, G., Piamrat, K., Alazab, M., Bhattacharya, S., Maddikunta, P. K. R. e Gadekallu, T. R. (2022). Federated learning for intrusion detection system: Concepts, challenges and future directions. Computer Communications, 195:346–361.

Andreoni Lopez, M., Mattos, D. M., Duarte, O. C. M. e Pujolle, G. (2019). Toward a monitoring and threat detection system based on stream processing as a virtual network function for big data. Concurrency and Computation: Practice and Experience, 31(20):e5344.

Azhagusundari, B., Thanamani, A. S. et al. (2013). Feature selection based on information gain. International Journal of Innovative Technology and Exploring Engineering (IJITEE), 2(2):18–21.

Brendan McMahan, H., Moore, E., Ramage, D., Hampson, S. e Agüera y Arcas, B. (2017). Communication-efficient learning of deep networks from decentralized data. Em Proceedings of the 20th International Conference on Artificial Intelligence and Statistics, AISTATS 2017, volume 54.

Cunha Neto, H. N., Hribar, J., Dusparic, I., Mattos, D. M. F. e Fernandes, N. C. (2023). A survey on securing federated learning: Analysis of applications, attacks, challenges, and trends. IEEE Access, 11:41928–41953.

Cunha Neto, H. N., Mattos, D. M. e Fernandes, N. C. (2021). Fedsa: Arrefecimento simulado federado para a aceleraçao da detecçao de intrusao em ambientes colaborativos. Em Anais do XXXIX Simpósio Brasileiro de Redes de Computadores e Sistemas Distribuídos, p. 280–293. SBC.

Cunha Neto, H. N., Mattos, D. M. F. e Fernandes, N. C. (2020). Privacidade do usuário em aprendizado colaborativo: Federated learning, da teoria à prática. Minicursos do Simpósio Brasileiro de Segurança de Informação e de Sistemas Computacionais - SBSeg, 20:142–195.

Kim, G., Kim, J. e Han, B. (2022). Communication-efficient federated learning with acceleration of global momentum. arXiv preprint arXiv:2201.03172.

Lai, F., Zhu, X., Madhyastha, H. V. e Chowdhury, M. (2021). Oort: Efficient federated learning via guided participant selection. Em 15th {USENIX} Symposium on Operating Systems Design and Implementation ({OSDI} 21), p. 19–35.

Lee, W. (2022). Reward-based participant selection for improving federated reinforcement learning. ICT Express.

Momma, M. e Bennett, K. P. (2002). A pattern search method for model selection of support vector regression. Em Proceedings of the 2002 SIAM International Conference on Data Mining, p. 261–274. SIAM.

Nishio, T. e Yonetani, R. (2019). Client selection for federated learning with heterogeneous resources in mobile edge. Em ICC 2019 2019 IEEE International Conference on Communications (ICC), p. 1–7.

Sharafaldin, I., Lashkari, A. H., Hakak, S. e Ghorbani, A. A. (2019). Developing realistic distributed denial of service (ddos) attack dataset and taxonomy. Em 2019 International Carnahan Conference on Security Technology (ICCST), p. 1–8.

Song, Z., Sun, H., Yang, H. H., Wang, X., Zhang, Y. e Quek, T. Q. (2021). Reputationbased federated learning for secure wireless networks. IEEE Internet of Things Journal, 9(2):1212–1226.

Wang, Y. e Kantarci, B. (2020). A novel reputation-aware client selection scheme for federated learning within mobile environments. Em 2020 IEEE 25th International Workshop on Computer Aided Modeling and Design of Communication Links and Networks (CAMAD), p. 1–6. IEEE.

Yoshida, N., Nishio, T., Morikura, M., Yamamoto, K. e Yonetani, R. (2020). Hybrid-fl for wireless networks: Cooperative learning mechanism using non-iid data. Em ICC 2020-2020 IEEE International Conference on Communications (ICC), p. 1–7. IEEE.