Design And Detection Of A Practical Quantum Random Number Generator

Random numbers have extensive applications in various fields,such as key generation and encryption algorithms in cryptography and information security,Monte Carlo methods and numerical calculations in computer simulations,and ensuring fairness and enjoyment in entertainment and games.With the development of quantum technology,we have entered a new era of science.Breakthroughs in quantum computing,quantum communication and quantum cryptography provide new possibilities for more efficient,secure and reliable information processing and transmission.In this process,high-quality random numbers play a crucial role.In quantum cryptography,random numbers are used for quantum key distribution to achieve unconditionally secure communication.At the same time,quantum technology also provides new ways to generate true random numbers.Quantum Random Number Generators(QRNGs)utilize quantum phenomena such as quantum entanglement and the random collapse of quantum states to generate random numbers with higher levels of randomness and unpredictability.This article mainly utilizes a development platform based on ZYNQ(FPGA+ARM)to implement a high-speed transmission scheme of quantum random numbers based on vacuum fluctuations.It also evaluates the performance of the generated random numbers using machine learning methods,covering the following aspects:(1)We introduced two popular development boards based on the ZYNQ series: Red Pitaya and ZCU111,and conducted Ethernet transmission testing on both boards.Next,a vacuum fluctuation-based QRNG was constructed,and the corresponding experimental system was successfully built.After data collection and post-processing,the generated random numbers exhibited high levels of randomness and unpredictability.In this process,we utilized the Red Pitaya development board to achieve high-speed data interaction between the processor and hardware resources.By utilizing a Gigabit Ethernet interface and employing a lightweight network protocol stack,we achieved high-speed transmission of random numbers and uploaded the generated random numbers to the computer.(2)We have established a Transformer-based machine learning model to quantify the randomness of the QRNG.Traditional test suites maily target the statistical characteristics of random number sequences,but they cannot detect inherent design flaws or attack vulnerabilities in random number generators.Our method is not only applicable to quantum random numbers with small data volumes,but it also quantifies their randomness.We compared the Transformer-based model with the LSTM-based model and found that the former has higher prediction accuracy than the latter,with its training time cost being only 47% of the latter’s.(3)Based on the size and layout requirements of the experimental device,a suitable sized chassis is designed through 3D modeling and then manufactured to encapsulate the entire experimental device,ensuring its stability,safety,and ease of portability and use.Fix key components such as experimental devices and development boards inside the chassis,and use appropriate brackets and screws to ensure their stability.To facilitate equipment maintenance and debugging,we have also reserved some openings and interfaces on the front panel to easily access the components when needed.