Continuous Identity Authentication Method Based on Channel State Information

The Internet of Things (IoT) technology is rapidly developing, and WiFi-based IoT devices utilize open wireless channels to transmit data, making them vulnerable to device spoofing attacks and communication data forgery. Achieving secure access for WiFi devices in resource-constrained environments and ensuring continuous identity authentication of accessing devices pose significant challenges. This paper proposes a method for continuous identity authentication of WiFi-based IoT devices based on Channel State Information (CSI), without altering the existing hardware structure of wireless IoT terminals. By extracting the channel state information from the wireless signals of the devices, a unique identity fingerprint is formed for each device. The channel state information of legitimate devices is collected to create a fingerprint database. The legitimate fingerprints are then input into a machine learning algorithm to train a classifier, which is used for continuous identity authentication of subsequent accessing devices. Experimental results demonstrate that the CSI-based fingerprint features can effectively and continuously identify WiFi-based IoT devices with high accuracy and good recognition performance.