Daily Arxiv

This page organizes papers related to artificial intelligence published around the world.
This page is summarized using Google Gemini and is operated on a non-profit basis.
The copyright of the paper belongs to the author and the relevant institution. When sharing, simply cite the source.

OptiFLIDS: Optimized Federated Learning for Energy-Efficient Intrusion Detection in IoT

Created by
  • Haebom

Author

Saida Elouardi, Mohammed Jouhari, Anas Motii

Outline

OptiFLIDS is a novel intrusion detection system (IDS) approach designed to secure critical IoT environments, such as smart homes and industrial systems. It utilizes federated learning (FL) to perform collaborative model training while preserving data privacy. OptiFLIDS applies pruning techniques during local training to reduce model complexity and energy consumption, and incorporates a custom aggregation method to handle differently pruned models due to non-IID data distributions. Experiments on three IoT IDS datasets—TON_IoT, X-IIoTID, and IDSIoT2024—show that OptiFLIDS improves energy efficiency while maintaining robust detection performance.

Takeaways, Limitations

•
Takeaways:
◦
Training intrusion detection models in IoT environments while maintaining privacy through federated learning.
◦
Reducing model complexity and energy consumption through model pruning.
◦
Implementing a custom aggregation method to handle non-IID data
◦
Development of an energy-efficient IDS model suitable for real-world IoT environments.
•
Limitations:
◦
The specific Limitations is not specified in the paper (but there may be Limitations of the federated learning itself, i.e. vulnerability to data imbalance, etc.)
◦
Lack of details on specific model pruning methods and custom aggregation methods (further research is needed).
👍
Made with Slashpage