Data Analysis Methods Preliminaries for a Photon-based Hardware Random Number Generator

Dmitriy Beznosko,Keith Driscoll,Fernando Guadarrama,Steven Mai,Nikolas Thornton

2024-05-15

Abstract:High quality random numbers are necessary in the modern world. Ranging from encryption keys in cyber security to models and simulations for scientific use: it's important that these random numbers are of high quality and quickly attainable. One common solution to the generation of random numbers is that of pseudo-random number generators, or PRNGs. PRNGs generate random numbers by first quantifying some unpredictable phenomena into a number or string and feeding it into an algorithm which yields numbers randomly based on that seed. Easy places to find seeds include the user's mouse movements or the machine's uptime. These are only pseudorandom, however, as if given the same seed twice, the PRNG would generate the same 'random' output. This is great for games like Minecraft, but not so great for cybersecurity encryption key generation. By using a hardware random number generator (HRNG), random numbers that are not susceptible to the flaws found in PRNGs can be attained at a high rate.

Cryptography and Security,Instrumentation and Detectors

What problem does this paper attempt to address?

This paper aims to solve the problems of high cost, slow speed and low quality in current hardware random number generators (HRNGs). Specifically, the author proposes a simple design based on low - photon absorption, using photomultiplier tubes (PMT) or silicon - based photodetectors to generate high - quality random numbers. This method can provide a large number of high - quality random numbers, which is suitable for computer applications requiring high security and simulation and modeling in scientific research.

The paper explores two different photon detection methods, one using photomultiplier tubes (PMT) and the other using multi - pixel photon counters (MPPC). It describes in detail the experimental setups of these two methods and discusses data processing methods, including the high - low method and the odd - even method. In addition, in order to evaluate the quality of the generated random numbers, the author proposes three test methods: arithmetic mean and standard deviation (AMSD) test, Monte Carlo pi - estimation (MCPE) test and the newly developed fraction - line symmetry (FLS) test.

Through these test methods, the author hopes to verify the effectiveness and reliability of the photon - based HRNG scheme. In particular, the FLS test is used to detect the quality of random numbers more meticulously in order to distinguish the effects of different processing methods. Eventually, the author expects that through these studies, it can be proved that the photon - based HRNG design can provide higher - quality and more - efficient random number generation than existing solutions.