[Paper Review] Ultra-fast real-time quantum random number generator with correlated measurement outcomes and rigorous security certification

Tobias Gehring, Cosmo Lupo+7 more|arXiv (Cornell University)|Dec 13, 2018

Chaos-based Image/Signal Encryption3 references4 citations

TL;DR

This paper presents a real-time quantum random number generator (QRNG) based on quadrature measurements of vacuum fluctuations, achieving 8 Gbit/s generation speed with rigorous security certification. It eliminates assumptions about classical adversaries and uncorrelated measurements by using a metrology-grade characterization of the homodyne detector, ensuring a lower bound on extractable randomness even under quantum attacks.

ABSTRACT

Quantum random number generators (QRNGs) promise perfectly unpredictable random numbers. However, the security certification of the random numbers in form of a stochastic model often introduces assumptions that are either hardly justified or indeed unnecessary. Two important examples are the restriction of an adversary to the classical regime as well as negligible correlations between consecutive measurement outcomes. Additionally, non-rigorous system characterization opens a security loophole. In this work we experimentally realize a QRNG that does not rely on the aforementioned assumptions and whose stochastic model is established by a rigorous -- metrological -- approach. Based on quadrature measurements of vacuum fluctuations, we demonstrate a real-time random number generation rate of 8 \,GBit/s. Our security certification approach offers a number of practical benefits and will therefore find widespread applications in quantum random number generators. In particular, our generated random numbers are well suited for today's conventional and quantum cryptographic solutions.

Motivation & Objective

• To address critical security loopholes in existing QRNGs, including assumptions about classical adversaries and uncorrelated measurements.
• To eliminate reliance on unverified or unnecessary assumptions in stochastic modeling of random number generation.
• To achieve real-time, high-speed random number generation (8 Gbit/s) with information-theoretic security certification.
• To develop a metrology-grade characterization of the homodyne detection system to rigorously bound parameter uncertainties and system failure probabilities.
• To demonstrate practical applicability of the QRNG in classical and quantum cryptographic systems with provable security.

Proposed method

• The QRNG uses quadrature measurements of vacuum fluctuations via a home-built homodyne detector based on a MAR-6 microwave amplifier and InGaAs photodiodes.
• A transfer function is experimentally characterized to model the system’s response, enabling a rigorous lower bound on vacuum fluctuation power.
• The system accounts for temporal correlations due to finite detection bandwidth by incorporating the frequency-dependent response R(ν) and visibility χ in the stochastic model.
• Quantum security is ensured by computing a lower bound on extractable randomness against a quantum-enabled adversary, using a conservative metrological approach.
• Random numbers are extracted in real-time using a Toeplitz randomness extractor implemented on an FPGA, with post-processing to remove residual correlations.
• The stochastic model is validated by comparing measured transfer functions with theoretical predictions, confirming the lower bound on vacuum fluctuations.

Experimental results

Research questions

• RQ1 Can a QRNG achieve real-time generation at 8 Gbit/s while rigorously certifying security without assuming classical adversaries?
• RQ2 How can temporal correlations in measurement outcomes be accounted for in the security model without throttling the generation rate?
• RQ3 What is the impact of detector imperfections and parameter uncertainties on the security certification of a QRNG?
• RQ4 Can a metrology-grade characterization of the homodyne system provide a lower bound on extractable randomness that remains valid under quantum attacks?
• RQ5 To what extent can the QRNG be practically applied in quantum and classical cryptography with provable security?

Key findings

• The QRNG achieves a real-time random number generation rate of 8 Gbit/s using quadrature measurements of vacuum fluctuations.
• The system demonstrates a rigorous security certification by establishing a metrology-grade stochastic model with bounded parameter uncertainties.
• The lower bound on extractable randomness is proven to be valid even against a quantum-enabled adversary, eliminating the need for classical-only adversary assumptions.
• Temporal correlations due to finite detection bandwidth are explicitly accounted for in the security model, ensuring robustness without rate reduction.
• The QRNG is shown to be suitable for quantum key distribution, with a security epsilon of approximately 10⁻³³ after 10 years of continuous operation.
• The transfer function characterization confirms that the measured vacuum fluctuation power is bounded below by the theoretical minimum, validating the stochastic model.