Radio Frequency Fingerprinting Enrollment for Proximity Integrated Circuit Cards

Radio frequency (RF) fingerprinting is essential for enhancing Physical layer (PHY) security, addressing key challenges in device identification. Feature extractors play a critical role by selecting the most relevant RF fingerprint features, enabling faster and more secure device enrollment. This streamlined enrollment process minimizes reliance on traditional key-based methods, optimizing the system for more efficient data transmission. In our proposed system, built on ISO/IEC 14443 NFC standards, a deep learning model trains the feature extractor to generalize to unseen Proximity Integrated Circuit Cards (PICCs). Focusing on the most relevant features, the system accurately classifies new devices with minimal enrollment data, significantly boosting security.