Assurance for quantum random number generators

数据是世界上最有价值的商品之一，它影响着每一个人、每一家公司、每一个政府、每一个地方。世界上大多数网络安全基础设施都是基于数字加密密钥的交换和使用。随机数生成器（rng）是这种现有基础设施和量子密钥分发等新技术的重要组成部分。量子随机数生成器（qrng）是利用自然物理过程中固有的随机性来创建随机数的设备，如果该过程是真正的量子，则确保每个设备都是唯一的，并且是量子技术的第一个实际实现之一。量子rng与其他传统伪rng的一个关键区别是，相同制造和制备的伪rng肯定会产生相同的随机序列，而qrng则不然，这对所有安全应用都至关重要。目前尚不存在对qrng产生的独特随机性进行权威评估的方法。该项目将解决这一需求，从而克服商业和工业开发的这一重要技术障碍，并最大限度地提高英国在该领域量子技术研究的回报。目前对随机数生成器（rng）的测试基于对其输出的数值分析，提供了关于输出随机性的统计特性的信息，但不能保证输出对其他人来说是未知的。对qrng进行更强的评估是可能的，因为除了数值分析以确保随机性之外，用于创建输出的物理过程可以建模并进行物理测试。评估该过程的“量子化”也评估了输出的隐私性。该项目将采用已经上市或接近市场原型的qrng，并实施这种评估方法。因此，它将为创建英国qrng评估过程提供专业知识和能力。

Data is one of the world's most valuable commodities -- affecting every person, every company, every government, everywhere. Most of the world's cybersecurity infrastructure is based on the exchange and use of digital cryptographic keys. Random number generators (RNGs) are essential components of this existing infrastructure, and newer technologies such as quantum key distribution. Quantum random number generators (QRNGs) are devices that utilise the inherent randomness in natural physical processes to create random numbers, assured unique to each device if the process is truly quantum, and are one of the first practical implementations of quantum technologies. A key differentiator of quantum RNGs over other conventional pseudo RNGs, crucial for all security applications, is that identically manufactured and prepared pseudo RNGs are certain to produce the same random sequences, while QRNGs are not.A method for providing authoritative assessment of the unique randomness produced by QRNGs does not currently exist. This project will address that need, thereby overcoming this important technological barrier to their commercial and industrial exploitation, and maximising UK return from quantum technology research in this field. Current tests for random number generators (RNGs), based on numerical analysis of their outputs, give information about the statistical properties of the output randomness but cannot assure that the output is unknown to others. Stronger assessment is possible for QRNGs, since in addition to numerical analysis to assure randomness, the physical process used to create the output can be modelled and physically tested. Assessing the "quantumness" of the process also assesses the privacy of the output.This project will take QRNGs that are either already on the market or near-market prototypes and implement this assessment approach. It will thereby provide the expertise and capability for creating a UK assessment process for QRNGs.