Identification of Cloned HF RFID Proximity Cards Based on RF Fingerprinting

Passively powered high frequency (HF) Radio Frequency Identification (RFID) proximity cards have been widely used in many fields. Meanwhile, a lot of security and privacy threats such as cloning, unauthorized tracking have arisen. Many security protocols have been proposed. However, most of them focused on security implication of the logical layer, little attention was paid to characteristics of the physical layer. In this work, we consider identification of HF RFID proximity cards by investigating electromagnetic characteristics of their physical layers. A new RF fingerprint extraction method is proposed based on normalized power spectral density of envelop of the card response. The most useful features for identification are found within 5 odd harmonics of digital transmission and the subcarrier modulation frequency. After evaluating our technique on a set of 300 RFID cards from 6 manufacturers, we can achieve Equal Error Rate (EER)=0 for all types by selecting appropriate thresholds. As the feature extraction can be implemented under normal working state of cards, it is practical and efficient to realize real-time identification of HF RFID proximity cards.