物联网设备检测方法、终端设备及存储介质
By transforming IoT device flows into device flow graphs and utilizing graph neural networks for feature learning, the problems of low accuracy in IoT device detection and inability to continuously learn new device types are solved, achieving efficient and accurate IoT device identification and continuous learning.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HUNAN UNIV
- Filing Date
- 2023-04-26
- Publication Date
- 2026-07-17
AI Technical Summary
Existing IoT device detection methods suffer from low detection accuracy and an inability to continuously learn new device types. In particular, those based on classic machine learning and deep learning models require a large amount of feature engineering and computing resources, and cannot quickly identify new device types.
The data packet information of the interaction between IoT devices and the gateway is transformed into a device flow graph. Feature learning is performed using a graph neural network NESGNN. Identification is performed through the vertex, edge features and subgraph structure of the device flow graph. Continuous learning is carried out by parameter sharing and sample replay.
It improves the accuracy of IoT device detection, reduces the complexity and computational cost of feature engineering, enables continuous learning of new device types, and reduces training time and resource consumption.
Smart Images
Figure CN116484257B_ABST