A Network Anomaly Monitoring Method Based on Edge Computing for CPS/IOT

In the large-scale IoT monitoring system, the cloud platform has been used as a remote data and control center. However, a large amount of data uploading and processing in the system brings great challenges to the cloud platform's bandwidth load and real-time feedback. Therefore, the paper proposed the framework of CPS/IoT monitoring system based on edge computing . On this basis, an anomaly detection method based on self-encoding neural network is proposed. The simulation data of the standardized pastures in Yunnan Province from 2017 to 2018 were used in the field. The simulation results show that the anomaly detection method based on self-encoding neural network can make full use of the spatial correlation of the collected data and accurately detect the anomaly data.