Intelligent Monitoring and Scheduling of Real-time Dynamic Data Based on Digital Twin

The Digital Twin (DT) is a key concept in the industrial Internet of Things (IoT) that creates virtual representations of physical objects through software definition. It maps the status, features, and evolution of physical entities in the digital space, providing a practical solution for capturing the dynamic and complex nature of industrial environments. This paper presents a proposed DT network architecture for managing devices in an information room. By integrating DT and edge computing, we establish an effective mapping between IoT equipment and network systems. Furthermore, we construct an edge twin network framework based on federated learning and introduce anomaly recognition models to make the twin network have predictive capability for abnormal data. Meanwhile, optimal strategy of communication resource allocation based on deep neural network model is determined to reduce the transmission delay, complete the prediction of the whole life cycle of the device, and realize real-time dynamic intelligent monitoring and scheduling.