SecFedNIDS: Robust defense for poisoning attack against federated learning-based network intrusion detection system

Zhao Zhang, Yong Zhang, Da Guo, Lei Yao, Zhao Li

Abstract: Highlights•Build a secure and robust federated learning-based network intrusion detection system.•We propose the model-level defensive mechanism based on poisoned model detection.•Based on important model parameter selection and online unsupervised poisoned model detection.•We propose a novel poisoned data detection method for the data-level defense.