Enhancing Intrusion Detection Robustness in Non-IID Federated Learning Systems

DOI

10.2991/978-94-6239-693-7_106How to use a DOI?

Keywords

Federated Learning; Intrusion Detection System; Non- IID Data; Robust Aggregation; Cybersecurity; Privacy Preservation; Machine Learning; NSL-KDD; CICIDS2017; Adversarial Robustness

Abstract

During the recent years, the blistering development of Internet of Things (IoT) and cyber-physical systems has required innovative, privacy-sensitive, and distributed cyber security systems. Federated Learning (FL) has become a paradigm shift model that allows several parties to cooperatively train a common model of intrusion detection without sharing raw data. Nevertheless, (non-independent and identically distributed) (non-IID) data among clients can dominate in real world network systems, resulting in worse convergence, worse generalization, adversarial weaknesses, etc. The study suggests a strong federated learning architecture designed to deal with intrusion detection when non-IID is in play, namely, it relies on adaptive aggregation, dynamic weighting and local normalisation to improve the robustness and stability of the models. The benchmark datasets used in the study, including NSL-KDD and CICIDS2017, are to be used in testing the performance of a proposed model. Findings demonstrate that it has a much higher detection and convergence consistency as well as client drift resilience compared to traditional FedAvg algorithms.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Kushagra Pal
AU  - Poornima Tyagi
AU  - Pradeep Kumar
PY  - 2026
DA  - 2026/06/16
TI  - Enhancing Intrusion Detection Robustness in Non-IID Federated Learning Systems
BT  - Proceedings of the International Conference on Intelligent Systems for a Sustainable Future (ISSF 2026)
PB  - Atlantis Press
SP  - 1096
EP  - 1111
SN  - 2589-4919
UR  - https://doi.org/10.2991/978-94-6239-693-7_106
DO  - 10.2991/978-94-6239-693-7_106
ID  - Pal2026
ER  -