Quantum-resilient optical links using micro-LEDs generated quantum random numbers and physical unclonable functions

Read full text
Authors: Tien Khee Ng, Heming Lin, Wenqing Niu, Tae‐Yong Park, Hang Lu, Zhican Zhou, Yating Wan, Boon S. Ooi
Cited by: 0
References: 12

Abstract

Abstract Secure communication in Internet-of-Things (IoT) systems typically rely on separate hardware elements for key generation, device authentication, and data transmission, complicating scalable integration. Here we demonstrate that micro-LEDs can integrate all three capabilities, reducing energy and footprint overheads and avoiding exposure to additional attack surfaces. Notably, fabrication-induced structural variations can generate stable, device-unique near-field emission patterns that serve as optoelectronic physical unclonable functions (PUFs), while spontaneous-emission noise can provide a quantum-origin entropy source for quantum random number generators (QRNGs). Through min-entropy evaluation and post-processing, a single micro-LED delivers a stable QRNG throughput of 13.25 Gb/s. Using a deep-neural-network-assisted extractor, each device yields reproducible 256-bit keys, with intra- and inter-device Hamming distances separated by more than an order of magnitude. Integrated into a post-quantum cryptographic framework, the micro-LED functions as a self-contained secure transmitter, enabling authenticated and encrypted free-space optical wireless communication at 1.97 Gb/s. In principle, the security mechanisms are wavelength-independent and extendable across material platforms. Together, these results establish a quantum-resilient photonic node that consolidates transmission, identity, and randomness at the edge device level, providing a scalable hardware building block for secure, low-power connectivity in future IoT networks.

Fields of Study & Subjects

Physical Unclonable Functions (PUFs) and Hardware Security Quantum Information and Cryptography Optical Wireless Communication Technologies

Publication details

Journal / source
Research Square
Publication date
4 March 2026
Language
EN
Document type
Preprint