Device Identification based on Network Traffic Fingerprint
Tianqi Wu, Xiaoming Zhou, Hongbin Wu, Daojuan Zhang, Yingjie Zhang, Xiao Han · 2023
The devices connected to the LAN have the security risk of being counterfeited and remotely controlled. To identify and understand the devices in the network in advance, the administrator can monitor and handle the abnormal devices in advance. At present, the device recognition method based on deep learning has some limitations. In this paper, the fingerprint information related to the device type is extracted from the network traffic, and the deep learning algorithm based on graph neural network is used to deeply mine the traffic feature hiding relationship between devices, and effectively identify various device types and device models. The effect of the model in this paper on the task of device identification of public data sets has been greatly improved.