Intrusion Detection in IoT: A Federated Learning Approach

The widespread adoption of Internet of Things (IoT) technologies has significantly expanded the attack surface of modern interconnected environments, thereby intensifying the need for scalable and privacy-aware Intrusion Detection Systems (IDS). Federated Learning (FL) has recently gained considerable attention as an effective collaborative learning framework that enables the training of IDS models without direct data sharing, thus enhancing both privacy protection and system efficiency. This survey systematically reviews ۲۰ recent studies on FL-based IoT intrusion detection, with a focus on class-incremental learning, hybrid edge–cloud architectures, secure aggregation mechanisms, feature reduction strategies, and multi-agent intelligence. The key methodologies, datasets, learning algorithms, and performance results reported in these studies are critically analyzed to identify recent progress as well as existing limitations. Finally, the paper outlines open research challenges and discusses future directions toward the practical deployment of FL-enabled IDS in real-world IoT environments.

کلیدواژه ها:

Federated Learning ، IoT Security ، Intrusion Detection ، Edge Computing ، Privacy Preservation