This paper presents an investigation of nucleotide sequencing based random number generators, refutation of naive approaches to this problem, and a novel random number generator design based on the characteristics of nucleotide sequencers such as the Oxford Nanopore Technologies (ONT) MinION. Common issues include misunderstanding the statistical properties of nucleotide sequences and the provenance of entropy observed in post-processed sequences extracted from such data. We identify that the use of sequences, expressed as base-pair (ATCG) sequences, for random number generation is not possible. The process by which such sequences are observed and reported by scientific instrumentation, provide a means by which entropy associated with nucleotide sequences (or more correctly the act of observing and recording them) can be observed. We report a novel method of extracting entropy from the process of reading nucleotide sequences, as opposed to the nucleotide sequences themselves. We overcome the limitations and inherent bias of nucleotide sequences, to provide a source of randomness decoupled from biological data and records. A novel random number generator drawing on entropy extracted from nucleotide sequencing is presented with validation of its performance and characteristics.
This paper presents an investigation of nucleotide sequencing based random number generators, refutation of naive approaches to this problem, and a novel random number generator design based on the characteristics of nucleotide sequencers such as the Oxford Nanopore Technologies (ONT) MinION. Common issues include misunderstanding the statistical properties of nucleotide sequences and the provenance of entropy observed in post-processed sequences extracted from such data. We identify that the use of sequences, expressed as base-pair (ATCG) sequences, for random number generation is not possible. The process by which such sequences are observed and reported by scientific instrumentation, provide a means by which entropy associated with nucleotide sequences (or more correctly the act of observing and recording them) can be observed. We report a novel method of extracting entropy from the process of reading nucleotide sequences, as opposed to the nucleotide sequences themselves. We overcome the limitations and inherent bias of nucleotide sequences, to provide a source of randomness decoupled from biological data and records. A novel random number generator drawing on entropy extracted from nucleotide sequencing is presented with validation of its performance and characteristics.
The volume of securely encrypted data transmission required by today's network complexity of people, transactions and interactions increases continuously. To guarantee security of encryption and decryption schemes for exchanging sensitive information, large volumes of true random numbers are required. Here we present a method to exploit the stochastic nature of chemistry by synthesizing DNA strands composed of random nucleotides. We compare three commercial random DNA syntheses giving a measure for robustness and synthesis distribution of nucleotides and show that using DNA for random number generation, we can obtain 7 million GB of randomness from one synthesis run, which can be read out using state-of-the-art sequencing technologies at rates of ca. 300 kB/s. Using the von Neumann algorithm for data compression, we remove bias introduced from human or technological sources and assess randomness using NIST's statistical test suite.
Random numbers (RNs) have applications in many areas: lottery games, gambling, computer simulation, and, most importantly, cryptography [N. Gisin et al., Rev. Mod. Phys. 74 (2002) 145]. In cryptography theory, the theoretical security of the system calls for high quality RNs. Therefore, developing methods for producing unpredictable RNs with adequate speed is an attractive topic. Early on, despite the lack of theoretical support, pseudo RNs generated by algorithmic methods performed well and satisfied reasonable statistical requirements. However, as implemented, those pseudorandom sequences were completely determined by mathematical formulas and initial seeds, which cannot introduce extra entropy or information. In these cases, “random” bits are generated that are not at all random. Physical random number generators (RNGs), which, in contrast to algorithmic methods, are based on unpredictable physical random phenomena, have attracted considerable research interest. However, the way that we extract random bits from those physical entropy sources has a large influence on the efficiency and performance of the system. In this manuscript, we will review and discuss several randomness extraction schemes that are based on radiation or photon arrival times. We analyze the robustness, post-processing requirements and, in particular, the extraction efficiency of those methods to aid in the construction of efficient, compact and robust physical RNG systems.
Random numbers play a significant role in fields like scientific simulations and cryptography. Here we describe a physical random number generator based on the intrinsic randomness of quantum mechanics. We design a simple method to produce and analyze random sequences starting with a source of correlated/entangled photon pairs as an entropy source. We generate random bits from the coincidence rates of photon pairs in 3 steps: (i) generate correlated/entangled photon pairs, (ii) convert the coincidence rates from decimal to binary, (iii) apply a randomness extraction procedure (post-processing). In this approach, we have tested the influence of entanglement on the entire generation process. In order to obtain a good-quality random sequance, we have used and compared two different extractors in the post-processing step.
Randomness is widely used in various fields, including encryption, statistical analysis, and numerical simulations. Randomness is also a fundamental resource in science and engineering. For such applications, we usually need to provide unbiased and independent random bits. This raises the issue of where to get these supposed random bits. Quantum random number generators (QRNGs) generate real random numbers based on the inherent randomness of quantum measurements. In practice, unfortunately, quantum randomness is inevitably mixed with classical randomness due to classical noise. Also, randomness is often correlated and biased. It is necessary to process the resulting raw bits sequence and convert them to good-quality output values that are as close to uniform distribution as possible. Random extractors are required for this. We analyze the randomness obtained by quantum random number generators as well as various examples of post-processing. We discuss the types of randomness extractors. Also, we use information-theoretically secure extractors and present improved novel randomness extraction.
Generators of random sequences used in high-end applications such as cryptography rely on entropy sources for their indeterminism. Physical processes governed by the laws of quantum mechanics are excellent sources of entropy available in nature. However, extracting enough entropy from such systems for generating truly random sequences is challenging while maintaining the feasibility of the extraction procedure for real-world applications. Here, we present a compact and an all-electronic van der Waals heterostructure-based device capable of detecting discrete charge fluctuations for extracting entropy from physical processes and use it for the generation of independent and identically distributed true random sequences. We extract a record-high value (>0.98 bits/bit) of min-entropy using the proposed scheme. We demonstrate an entropy generation rate tunable over multiple orders of magnitude and show the persistence of the underlying physical process for temperatures ranging from cryogenic to ambient conditions. We verify the random nature of the generated sequences using tests such as NIST SP 800-90B standard and other statistical measures and verify the suitability of our random sequence for cryptographic applications using the NIST SP 800-22 standard. The generated random sequences are then used in implementing various randomized algorithms without any preconditioning steps.
Random number generators are indispensable for a multi- tude of tasks; from electronic games to secure communications. Most gen- erators have been made either in software or determinist hardwired de- vices such as the Linear-Feedback-Shift-Registers; while gaining in costs or speed, the “random” sequences generated are actually deterministic, obeying clear generating algorithms, despite all randomness appearance of their outputs. From the other side, Nature presents a multitude of sources of true randomness that can be explored. Commercial random generators exist based on physical processes as the source of randomness. Difficulties are always present to extract Nature’s randomness. This paper presents guidelines for construction of a fast (telecommunication speed) Physical Random Number Generator. It discusses the fundamental physical ele- ments involved, technicalities of signal recording and its limitations, and the final bit extraction. The need for randomness tests is emphasized and the impossibility of guaranteeing true randomness of a finite sequence is discussed.