Security of Wireless Devices using Biological-Inspired RF Fingerprinting Technique
Rehman, Saeed; Alam, Shafiq; Ardekani, Iman
Citation:Rehman, S., Alam, S. and Ardekani, I. T. (2014). Security of Wireless Devices using Biological-Inspired RF Fingerprinting Technique. In Biologically-Inspired Techniques for Knowledge Discovery and Data Mining. IGI Global(Eds.), (p..311-330). NOTE: PARTIAL EXTRACT FROM CHAPTER
Permanent link to Research Bank record:http://hdl.handle.net/10652/2958
Radio Frequency (RF) fingerprinting is a security mechanism inspired by biological fingerprint identification systems. RF fingerprinting is proposed as a means of providing an additional layer of security for wireless devices. RF fingerprinting classification is performed by selecting an “unknown” signal from the pool, generating its RF fingerprint, and using a classifier to correlate the received RF fingerprint witheach profile RF fingerprint stored in the database. Unlike a human biological fingerprint, RF fingerprint of a wireless device changes with the received Signal to Noise Ratio (SNR) and varies due to mobility of the transmitter/receiver and environment. The variations in the features of RF fingerprints affect the classification results of the RF fingerprinting. This chapter evaluates the performance of the KNN and neural network classification for varying SNR. Performance analysis is performed for three scenarios that correspond to the situation, when either transmitter or receiver is mobile, and SNR changes from low to high or vice versa.