RF Fingerprint Identification for 5G Mobile Device Based on Transient Features

In this paper, we propose an radio frequency (RF) fingerprint extraction method based on transient features for 5G mobile phones. From the perspective of network access security, the RF fingerprint of the physical random access channel (PRACH) signal has been studied. We find that the same PRACH preamble from the same device has the same transient features, while different PRACH preambles from the same device have different transient features. However, with a sufficiently large training set, classification can be achieved with an accuracy of 75.8% in the mixed-preamble case. 4 mobile phones are used in the experiment and a complete network access system has been built through OpenAirInterface.