Skip to content
Conference

Privacy-Preserving Intrusion Detection using Federated Learning and Tabular Generative Adversarial Networks

Oct 2026 · International Conference on Communication and Electronics Systems · pp. 1794-1800 · 0 citations · 22 references

Abstract

In this research, a solution towards private data collaboration through Federated Learning (FL), Conditional Tabular Generative Adversarial Networks (CTGAN), and Differential Privacy (DP) is investigated. This solution allows organisations to collaborate on model training without exchanging their original network traffic logs. We simulated the generation of real network traffic logs with local CTGAN models on each client and evaluated the results for feature distribution similarity, privacy leakage, and intrusion detection utility. We found a strong preservation of feature distributions and correlations in synthetic data. We showed that these generated logs resist empherical memorisation and identity disclosure, as evidenced by low nearest-neighbor overlap and very few duplicate records. When integrated into a federated IDS with DP, the pipeline provided a high level of privacy and utility when compared against a non-synthetic solution. An empirical distinguishability test confirmed that the logs are still separable from genuine network traffic. However, the pipeline has successfully balanced utility with privacy. These findings suggest that we can achieve powerful sharing of Cyber Intelligence, with synthesized logs acting as proxies for logs in federated IDS.

View source

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.