An Improved Sensor Anomaly Detection Method in IoT System using Federated Learning

The industrial sensor has emerged as a critical device to monitor environment condition in the manufacturing system. However, abnormal behaviors of these smart sensor may indicate some failure or potential risk during system operation, thereby increasing high availability of the entire manufacturing process. Data collected from many edge devices for detecting failure contain private data of different enterprises which is challenging current detection approaches as user privacy has attracted more concerns. Moreover, detecting anomalies in the centralized system is often more time consuming due to the response time. To overcome these issues, we proposed an anomaly detection method using a clustering federated learning framework with a long short-term memory (LSTM) to improve model performance in term of accuracy, scalability and more secure.