On Feasibility Of Fingerprinting Wireless Sensor Nodes Using Physical Properties

Abstract

Fingerprinting wireless devices using physical properties has been recently suggested as an alternative for device identification and authentication. It has been found that the clock skew caused by the frequency discrepancy of the quartz crystals in different devices can be used as a reliable source for fingerprinting. Researchers have studied the application of the clock skew-based fingerprinting in sensor networks and claimed that it can detect fake identities, wormholes, and node replicas. However, the study in this paper draws a completely opposite conclusion, i.e., the clock skew of sensor nodes can be easily forged by adversaries to evade the detection. This paper then studies the feasibility of using the distribution of signal power in space to fingerprint sensor nodes. The result shows that a sensor node's signal power distribution in space is not only reliable for being used as a source for fingerprinting but also very hard to forge. Finally, the paper discusses the application of using signal power distribution for detecting various attacks as well as the limitations and open problems.