EdgeDetect: Importance-Aware Gradient Compression with Homomorphic Aggregation for Federated Intrusion Detection

EdgeDetect is a federated intrusion detection system for 6G-IoT that enables distributed learning without centralizing raw data.

It uses median-based binarization to compress gradients and Paillier homomorphic encryption to protect privacy during aggregation.

Tests on CIC-IDS2017 and Raspberry Pi 4 showed high detection accuracy, strong robustness, and major communication savings.

The work tackles two key barriers to federated security models at once: heavy communication cost and gradient leakage.