A method for secure aggregation and real-time anomaly detection of IoT data based on federated learning

Abstract

The present invention discloses an IoT data security aggregation and real-time anomaly detection method based on federated learning, which relates to the field of IoT data processing and information security technology.This method integrates the collaborative capabilities of IoT terminal devices, edge nodes, and cloud servers through a layered federated learning architecture, achieving distributed secure aggregation of data and efficient anomaly detection.The present invention effectively solves the problems of privacy leakage, high centralized processing delay, and high false alarm rate in traditional IoT data aggregation. On the basis of ensuring data 10 security and privacy, it improves the real-time and reliability of anomaly detection and is suitable for IoT data security governance in multiple scenarios such as smart homes, industrial IoT, and intelligent transportation.