Dataset for Radio-Frequency Fingerprinting Millimeter-Wave Commercial Off-the-Shelf Cellular User Equipment

This data is associated with the manuscript titled Radio-Frequency Fingerprinting Millimeter-Wave Commercial Off-the-Shelf Cellular User Equipment. The abstract of this manuscript is: We present a measurement test bed and algorithm for non-destructive, millimeter wave over-the-air fingerprinting of commercial off-the-shelf cellular user equipment (UE) at the model level. This test bed has been configured to repeatably collect radiated fields from UEs in a 5G mmWave scenario. We evaluate the performance of a classification algorithm that uses linear discriminant analysis on two types of measured waveforms from cellular user equipment under test. One of the waveforms is the user equipment transmitted symbol constellation vector, which is complex-valued. The classification algorithm uses the symbol constellation waveforms collected from the UEs to generate a unique fingerprint of the test UE. The other waveform is the filtered uplink spectrum waveform from the UEs. We evaluate the performance of our test bed and algorithm by analyzing the temporal stability of the data sets for the purposes of fingerprinting and quantifying the classification performance of the test bed and algorithm.