A Research of I/Q Imbalance based RF Fingerprint Identification with LTE-RACH Signals

Abstract

Radio frequency fingerprint (RFF) based wireless device identification has been extensively researched. Long-Term Evolution (LTE) and 5th Generation (5G) mobile communication systems have similar random access channel (RACH) signals, which could be employed for RFF based device identifications. This paper proposes a practical I/Q imbalance-based RFF extraction method using LTE-RACH signals. We initially evaluate the performance of the proposed method by simulating LTE-RACH signals with RFF model. The relationship between I/Q imbalance characteristics and LTE-RACH signal format is explored. Secondly, we build an experimental system including 12 LTE mobile phones from 3 brands and 6 Universal Software Radio Peripheral (USRP) devices from 2 brands. The I/Q imbalance features are extracted for classifications. The experimental results show that the I/Q imbalance feature is distinguishable among the devices produced by different manufacturers, especially the commercial LTE mobile phones and experimental USRP devices. The classification accuracy rate of LTE mobile phones and USRP devices can reach 96.01%, and the accuracy rate of LTE mobile phones from different brands is 89.80%.