Physical unclonable functions (PUFs) have emerged as a promising encryption technology, utilizing intrinsic physical identifiers that offer enhanced security and tamper resistance. Multi-level PUFs boost system complexity, thereby improving system reliability and fault tolerance. However, crosstalk-free multi-level PUFs remain a persistent challenge. In this study, a hierarchical PUF system that harnesses the spontaneous phase separation of silk fibroin /PVA blend and the random distribution of silicon-vacancy diamonds within the blend is presented. The thermodynamic instability of phase separation and inherent unpredictability of diamond dispersion gives rise to intricate random patterns at two distinct scales, enabling time-efficient hierarchical authentication for cryptographic keys. These patterns are complementary yet independent, inherently resistant to replication and damage thus affording robust security and reliability to the proposed system. Furthermore, customized authentication algorithms are constructed: visual PUFs authentication utilizes neural network combined structural similarity index measure, while spectral PUFs authentication employs Hamming distance and cross-correlation bit operation. This hierarchical PUF system attains a high recognition rate without interscale crosstalk. Additionally, the coding capacity is exponentially enhanced using M-ary encoding to reinforce multi-level encryption. Hierarchical PUFs hold significant potential for immediate application, offering unprecedented data protection and cryptographic key authentication capabilities.
Keywords: anticounterfeit; diamond; phase separation; physical unclonable function; silicon vacancy.
© 2024 The Authors. Advanced Science published by Wiley‐VCH GmbH.