Empowering Privacy-Preserving Anomaly Detection: A New Framework for Cybersecurity

A study on Information Technology and Data Encoding and Encryption has identified a significant challenge in protecting against cyber threats. The researchers from the University of Ha'il, in collaboration with Hassan II University of Casablanca, have proposed a novel framework, Empowering Privacy-Preserving Anomaly Detection (EPAD), to address this challenge. EPAD integrates a lightweight homomorphic encryption library and federated learning, providing a robust and scalable solution for real-time intrusion detection.

Key Takeaways:

• EPAD addresses the limitations of traditional intrusion detection systems (IDS) by ensuring data confidentiality through encrypted computations performed locally at each client.
• The framework integrates multiple real-world and benchmark datasets, including BOT-IoT, NSL-KDD, ToN-IoT, and UNSW-NB15, to enable practical deployment in resource-constrained environments such as IoT and edge computing.
• The researchers optimized the homomorphic encryption library to reduce computational overhead while preserving data privacy.
• EPAD ensures data confidentiality by only sharing encrypted model updates with the central server.
• Experimental results demonstrate that EPAD achieves an accuracy of 98.21%, a precision of 99.14%, a recall of 99.01%, and an F1-score of 97.13%.

Statistics:

• The execution time of EPAD is reduced to 0.2 seconds.
• The accuracy of EPAD is 98.21%.
• The precision of EPAD is 99.14%.
• The recall of EPAD is 99.01%.
• The F1-score of EPAD is 97.13%.

Sources:

• Epad: Empowering Privacy-preserving Anomaly Detection Through Homomorphic Encryption and Machine Learning. Cluster Computing, 2025;28(10).
• Springer. Cluster Computing. www.springerlink.com/content/1386-7857/
• University of Ha'il. Appl Coll. Dept. of Computer Sciences. Hail 55424, Saudi Arabia.
• Hassan II University of Casablanca.