Tri-Type QRNG Provides High-Speed Randomness For Quantum
Quantum Computing Innovation: Tri-Type QRNG Provides On-Demand Randomness at Record Speeds Because the digital world needs genuinely random numbers for cryptography, Monte Carlo simulations, and advanced artificial intelligence, Quantum Random Number Generation (QRNG) has grown constantly. Because their mathematical algorithm and seed can be predicted, conventional pseudorandom number generators (PRNGs) pose major security risks. QRNGs use vacuum noise or photon statistics to produce random numbers.
A key shortcoming of most older QRNGs is that they only produce one type of random number distribution. Fluid mechanics and aerodynamics use Rayleigh distributed random numbers, financial modelling and climate simulations use Gaussian random numbers, and machine learning and cryptography need uniformly distributed random numbers. Until recently, coping with this diversity of applications required deploying numerous hardware systems or converting one random distribution to another, which sacrificed many secure bits and risked numerical mistakes. Researchers have constructed and proved the first tri-type QRNG system, which can measure quantum and process the results to produce uniform, Gaussian, and Rayleigh random bits on demand using the same hardware. Multi-Distribution Output with Quantum Vacuum Noise
Dual-quadrature homodyne detection monitors quantum vacuum noise, giving the tri-type QRNG versatility. Unlike earlier single-type QRNGs, this approach measures both quadrature simultaneously. The raw data from these measures naturally forms all three essential distributions: By measuring the vacuum state's quadrature using homodyne detection, Gaussian-distributed raw random numbers can be obtained.
Uniform and Rayleigh Distributions: Phase space charts can yield uniform and Rayleigh distributed digital random numbers. Amplitude is Rayleigh, but phase angle is uniform. Importantly, an FPGA board's electrical postprocessing step handles all the functionality to switch between the three distribution methods, ensuring that the quantum bits' high production rate is unaffected. Speed and Accessibility Standards The experimental device generates over 60 Gbits/s (Gbps) of raw bits, demonstrating its speed. The needed randomness extraction process eliminates classical side channel information and ensures quantum randomness, resulting in high safe bit rates: Uniform Random Numbers: Over 42 Gbps safe bit rate. This consistent random number passed NIST, Dieharder, and other rigorous statistical tests. Researchers suggested optimising the high-speed, low-noise anti-aliasing filter to raise the uniform extraction rate to over 68.25 Gbps. Secure bit rate extraction over 14 Gbps with Gaussian Random Numbers. A better extractor and filter could boost this pace to 22 Gbps. This device's hardware uses a 1550 nm fiber-coupled CW laser. Split laser light disrupts quantum vacuum state. Two homodyne detectors collect measurement signals, which are digitally processed by an AMD ZCU111 RFSoC development board. Additionally, technology is becoming more accessible in real life. Users can access the Cisco Quantum Random Number Service with statistically confirmed numbers uploaded to the Cisco Cloud after real-time testing. Security bits are continuously created on the FPGA board. Extraction and Statistical Integrity Challenge A robust randomness extractor must be applied to noisy raw quantum measurement results for great security. The well-known Toeplitz extractor, based on the universal hashing function, ensures net randomness extraction even with a reusable seed for uniform random numbers.
High-quality Rayleigh and Gaussian numbers are challenging to post-process. Gaussian Extraction: The system uses a Wallace-based modified recursive approach. The Most Significant Bits (MSBs) are selected using entropy-based truncation to reduce classical noise, which influences the LSBs. Even if the extracted Gaussian data passed Goodness-of-Fit (GoF) tests, the approach does not necessarily preserve randomness. This limits its usage in high-security encryption, but it is nevertheless helpful and fast for modelling, simulation, and probabilistic applications that require statistically correct Gaussian distributions. No comprehensive quantum randomness extractor for the Rayleigh distribution exists. Researchers are studying classical filters like the Savitzky-Golay filter to denoize raw Rayleigh bits. The tri-type QRNG's successful demonstration marks a milestone in quantum randomness generation and meets the multiple demands of modern security and computing applications with unprecedented speed and variety from a single platform.
