Random Number Generators in Advanced Optical Experiments: A Comparative Analysis of Semiclassical, Quantum, and Hybrid Architectures

TL;DR

This paper compares optical random number generators, analyzes their trade-offs, and introduces a hybrid architecture that combines sources to produce high-quality, high-rate random sequences for advanced experiments.

Contribution

It presents a novel hybrid optical random number generator architecture that improves quality and rate by combining different photon sources.

Findings

Hybrid architecture enables high-quality RNSs at increased rates.

Raw sequences from the hybrid source can surpass processed sequences in randomness.

Trade-offs between laser-based and heralded single-photon sources are quantitatively analyzed.

Abstract

Random numbers sequences (RNSs) play a vital role in various scientific and engineering applications. They are critical to the integrity of classical and quantum cryptography, the accuracy of mathematical modeling and Monte Carlo simulations, and the core mechanics of applications in fields as diverse as gambling and statistical sampling. While the primary criteria for RNSs sources are their quality and generation rate, their integration into experimental designs is equally significant for many fundamental physical tests and applications. This work presents a comparative analysis of optical random number generation architectures, which can be seamlessly included into various advanced classical and quantum optical experimental schemes. In particular, we evaluate the trade-off between the high generation rate of an attenuated laser (a quasi-single-photon source) and the superior statistical quality of a heralded single-photon source operating at a much lower frequency. To overcome the limitations of each individual source, we propose and examine a novel hybrid architecture that utilizes their mixed radiation, enabling the generation of highquality RNSs at an enhanced rate. Furthermore, we demonstrate that the raw sequences generated by such a source can not only exhibit but, in some cases, even surpass the degree of randomness achieved by sequences processed through powerful randomness extractors.