The pursuit of eco-sustainable materials for next-generation radio-frequency and sensing devices is fostering the development of antennas and tags that are not only recyclable but also capable of ensuring physical-level authenticity. Among these, Laser-Induced Graphene (LIG) enables the direct fabrication of conductive structures on polymeric substrates. Its intrinsic fabrication variability introduces microscopic randomness that can be exploited to realize Physical Unclonable Functions (PUFs) for secure authentication. This work proposes a near-field interrogation and processing framework for LIG-based UHF-RFID antennas, enabling the extraction of digital cryptographic keys from the electromagnetic backscattered response. A dedicated signal-processing and binarization pipeline converts analog amplitude and I/Q responses into statistically balanced binary databases. An experimental campaign on 30 devices, tested with two reader antennas and two orthogonal orientations, demonstrates that the proposed method can generate up to 76 independent bits per key, with reliability exceeding 0.8 and decidability values up to 8, ensuring both repeatability and device-level distinctiveness. By varying frequency, power, and orientation, multiple uncorrelated key families can be generated.
Nanni et al. (Thu,) studied this question.