FedNIDS: A Federated Learning Framework for Packet-Based Network Intrusion Detection System
Digital Threats: Research and Practice · 31.03.2025 · 6 (1) · 1-23
DOI: 10.1145/3696012 ↗Kurz gefasst
Fachartikel in der Zeitschrift Digital Threats: Research and Practice, online erschienen am 14.02.2025. Verfasst von Quoc H. Nguyen und 4 weiteren Autor:innen. Die Arbeit wurde bisher 30-mal zitiert (Crossref) und nennt 60 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: DEVCOM Army Research Laboratory, U.S. Military Academy.
Abstract (Originalsprache)
Network intrusion detection systems (NIDS) play a critical role in discerning between benign and malicious network traffic. Deep neural networks (DNNs), anchored on large and diverse datasets, exhibit promise in enhancing the detection accuracy of NIDS by capturing intricate network traffic patterns. However, safeguarding distributed computer networks against emerging cyber threats is increasingly challenging. Despite the abundance of diverse network data, decentralization persists due to data privacy and security concerns. This confers an asymmetric advantage to adversaries, as distributed networks face the formidable task of securely and efficiently sharing non-independently and identically distributed data to counter cyber-attacks. To address this, we propose the federated NIDS ( FedNIDS ), a novel two-stage framework that combines the power of federated learning and DNNs. It aims to enhance the detection accuracy of known attacks, robustness and resilience to novel attack patterns, and privacy preservation, using packet-level granular data. In the first stage, a global DNN model is collaboratively trained on distributed data, and the second stage adapts it to novel attack patterns. Our experiments on real-world intrusion datasets demonstrate the effectiveness of FedNIDS by achieving an average F1 score of 0.97 across distributed networks and quickly disseminating novel attack information within four rounds of communication.
Förderung
- DEVCOM Army Research Laboratory
- U.S. Military Academy
Zitieren
APA
Nguyen, Q. H., Hore, S., Shah, A., Le, T., & Bastian, N. D. (2025). FedNIDS: A Federated Learning Framework for Packet-Based Network Intrusion Detection System. Digital Threats: Research and Practice, 6(1), 1-23. https://doi.org/10.1145/3696012
Vancouver
Nguyen QH, Hore S, Shah A, Le T, Bastian ND. FedNIDS: A Federated Learning Framework for Packet-Based Network Intrusion Detection System. Digital Threats: Research and Practice. 2025;6(1):1-23. doi:10.1145/3696012
BibTeX
@article{Nguyen2025,
author = {Nguyen, Quoc H. and Hore, Soumyadeep and Shah, Ankit and Le, Trung and Bastian, Nathaniel D.},
title = {FedNIDS: A Federated Learning Framework for Packet-Based Network Intrusion Detection System},
journal = {Digital Threats: Research and Practice},
year = {2025},
volume = {6},
number = {1},
pages = {1--23},
doi = {10.1145/3696012}
}
EUVIMED