All-silicon multidimensionally-encoded optical physical unclonable functions for integrated circuit anti-counterfeiting

• Published in: Nature communications 2024-04, Vol.15 (1), p.3203-11, Article 3203
• Main Authors: Wang, Kun, Shi, Jianwei, Lai, Wenxuan, He, Qiang, Xu, Jun, Ni, Zhenyi, Liu, Xinfeng, Pi, Xiaodong, Yang, Deren
• Format: Article
• Language: English
• Subjects: 639/624/1075; 639/925/357/1015; 639/925/357/1017; Authentication; CMOS; Coding; Counterfeiting; Cybersecurity; Entropy; Entropy (Information theory); Erbium; Fabrication; Humanities and Social Sciences; Incompatibility; Integrated circuits; Internet of Things; multidisciplinary; Pixels; Quantum dots; Rejection rate; Science; Science (multidisciplinary); Silicon
• ISSN: 2041-1723; 2041-1723
• DOI: 10.1038/s41467-024-47479-y

Integrated circuit anti-counterfeiting based on optical physical unclonable functions (PUFs) plays a crucial role in guaranteeing secure identification and authentication for Internet of Things (IoT) devices. While considerable efforts have been devoted to exploring optical PUFs, two critical challenges remain: incompatibility with the complementary metal-oxide-semiconductor (CMOS) technology and limited information entropy. Here, we demonstrate all-silicon multidimensionally-encoded optical PUFs fabricated by integrating silicon (Si) metasurface and erbium-doped Si quantum dots (Er-Si QDs) with a CMOS-compatible procedure. Five in-situ optical responses have been manifested within a single pixel, rendering an ultrahigh information entropy of 2.32 bits/pixel. The position-dependent optical responses originate from the position-dependent radiation field and Purcell effect. Our evaluation highlights their potential in IoT security through advanced metrics like bit uniformity, similarity, intra- and inter-Hamming distance, false-acceptance and rejection rates, and encoding capacity. We finally demonstrate the implementation of efficient lightweight mutual authentication protocols for IoT applications by using the all-Si multidimensionally-encoded optical PUFs. The researchers introduce an all-silicon optical PUF that enhances IoT device security through CMOS-compatible fabrication, showcasing five unique optical responses per pixel for advanced authentication and high information entropy.