RF fingerprinting, RF front end, Classification, 

In wireless networks, physical layer security would be complementary when higher level software based security approaches are inadequate. One of the physical layer methods is Radio Frequency (RF) fingerprinting. In RF fingerprinting method, data acquisition phase plays a critical role in precisely capturing signals for extracting the fingerprints. In this thesis, a low-cost modular RF front-end designed with commercial of the shelf (COTS) is presented. In order to design this RF front end, computer based design tool is used. Three different data acquisition methods including proposed RF front end usage is also presented. Moreover, assessment of the RF front end for the RF fingerprinting performance with respect to three different data sets are presented. Transient signal detection, feature extraction and classification algorithms are implemented on these data sets. To do classification support vector machine (SVM) and neural networks (NN) are used. Results of the study show that the designed RF front end would work well in RF fingerprinting, and accurate classification of BT devices.