Random numbers play a crucial role in many applications, in particular in cryptography, games and computer simulations. It is of fundamental importance that they are truly unpredictable, as any deviation may adversely affect modelling or jeopardise security. Notably, recent breaches of cryptographic protocols have exploited weaknesses in the random number generation.
In this context, schemes exploiting the inherent randomness of quantum physics have been investigated and quantum random number generators (QRNG) were commercially available before the project started. However, QRANGE has pushed the QRNG technology even further, allowing for a wide range of commercial applications.
We targeted the development of three different prototypes, which are cheaper, faster and more secure than existing devices: a) A fully integrated low-cost QRNG based on standard CMOS technology with a target cost of the order of 1€ for large-scale production. b) A high-speed phase-diffusion scheme (based on the interference of laser pulses with random phases) with bit rates of up to 10Gb/s. c) A self-testing QRNG, which allows for a continuous estimation of the generated entropy, with few assumptions on the devices. Moreover, we made a considerable theoretical effort for modelling the devices, designing efficient randomness extractors and studying new semi device-independent concepts. Finally, we worked with the relevant institutions towards a certification scheme of QRNG devices compliant with the highest security standards.
QRANGE addressed many key points of the Quantum Technologies Flagship and is aligned with its objectives, taking quantum technologies from the laboratory to industry, with concrete prototype applications and products.
The conclusions are the following:
• QRNGs can be integrated in a wide spectrum of use cases, taking advantage of testable security and high random bit rates.
• The security of QRNGs is based on a physical modelling based on first principles. The prototypes developed are basically certification-ready by design and exhibit high-quality entropy sources.
• The ITU-T recommendation X.1702 is the first standard world-wide on QRNG entropy source architecture.
• Optical QRNGs show a significant potential for cost and size reductions, which will give the technology a significant share of the RNG market within this decade.
• Self-testing QRNG concepts allow to prove a certain level of entropy generation in real time, a method which is not accessible to non-quantum RNGs.
• Fast QRNGs beyond GB/s bit generation rate are technically feasible, show potential for cost and size reduction, and constitute important technology for HPC and QKD use cases.
• Theoretical improvements of the entropy source architecture and design are still possible. They increase the security and simplify the implementation by relaxing the constraints in the security proofs.
• Randomness extraction can be optimised by adapting the existing methods to quantum entropy sources.
