Using Radio Device Fingerprinting for the Detection of Impersonation and Sybil Attacks in Wireless Networks

Abstract

This paper describes an intrusion detection system to identify impersonation attacks and Sybil attacks in wireless networks. The detection system uses radio device fingerprinting and has experimental performance comparable with existing intrusion detection methods. The detection systems for Sybil attacks have not been widely investigated to date, and this contribution of the paper is novel. The paper also derives analytical formulae relating metrics of the fingerprinting classification procedure with the metrics of the intrusion detection system. The formulae can be used to guide the selection of the fingerprinting classification method, knowing the desired performance of the detection system. The use of radio device fingerprinting simplifies the task of securing a wireless ad-hoc network.