Continuous Authentication for the Internet of Things Using Channel State Information

In the centralized Internet of Things (IoT) architecture there exists gateways, which gather sensed data from sensor nodes frequently. Due to numerous security threats and attacks in wireless sensor networks (WSNs), how to achieve device -to -device authentication is a fundamental security issue. In some IoT environments such as environmental monitoring systems, sensor nodes need to transmit sensed data to gateways periodically. Previous authentication schemes make sensor nodes and gateways authenticate each other frequently in the beginning of each transmission session and affect the real-time nature of sensed data. In this paper, we explore a lightweight device fingerprint extraction and verification algorithm using Channel State Information (CSI). We propose a CSI -based continuous authentication scheme for the Internet of Things. The proposed scheme can achieve continuous authentication for the target IoT device and finish each authentication process within 1.5 ms. Our experiments show that the proposed scheme can resist identity-based attacks and has better performance than existing continuous authentication schemes for the Internet of Things.