Towards on-chip demonstration of a high-dimensional quantum random number generator

Data protection and confidentiality have become a serious concern in today’s world. Their security is guaranteed by cryptographic protocols, which heavily rely on random numbers as a measure against predictability. Classically, randomness is generated via complex but deterministic algorithms, which are vulnerable to attacks. Quantum random number generators (QRNGs) have emerged as a promising solution, as they provide true random numbers based on the intrinsic non-deterministic nature of quantum mechanics. However, critical challenges for QRNGs are the certification and quantification of their genuine randomness, especially in the presence of untrusted devices, and their compactness for systematic deployment. In this feasibility study, to face these challenges, we propose to use a silicon-photonic platform, leveraging on the concept of quantum contextuality for a semi-device independent generator. In particular, we use Klyachko-Can-Binicioglu-Shumovsky (KCBS) inequality to assess a fundamental property of quantum measurements: that their outcomes depend on the specific measurement context.