Analysis of Wireless Signature Feature Sets for Commercial IoT Devices : Invited Presentation

Abstract

Advances in technology have led to an increase the number and type of electronic devices interconnected through the internet, or Internet of Things (IoT) devices. These devices are available commercially and are often used for applications in home environments, office buildings, medical facilities, and others. A common protocol used in IoT devices is the Institute of Electrical and Electronics Engineers (IEEE) 802.15.4 protocol. The vulnerabilities of the standard have led to numerous attacks on devices that follow this protocol. RF fingerprinting is a technique used to authenticate and verify devices in an environment to determine if they are either authorized or rogue to add a level of security at the physical layer. The fingerprints are comprised of statistical features, such as variance, skewness, and kurtosis, extracted from instantaneous RF signal characteristics. Previous RF fingerprinting work obtained features from generic ZigBee modules. This work aims to examine signal features of commercial IoT devices that adhere to the ZigBee protocol. The analysis will highlight correlations, if any, and differences between device vendor and IoT device type.