Quantum random number generator for industrial applications

Carlos Abellan Sanchez

Valerio Pruneri

Randomness is one of the most intriguing, inspiring and debated topics throughout history. It is a concept that arises when we ask ourselves about our own existence: Why are we the way we are? Do we have free will? Is evolution the result of chance? And at the same time, when we try to understand our relationship with the universe itself. Why are we here? When or how did it all begin? Is the universe a deterministic machine or is there room for chance? Surprisingly, randomness also plays a crucial role in the information and technology era. Random numbers are used in communication protocols such as Ethernet, in classification and processing algorithms such as Page Rank. It is also used in Monte Carlo methods, within the fields of physics, biology, chemistry, finance or mathematics. But the most iconic application for random numbers is found in the field of cryptography or cybersecurity where random numbers generate cryptographic keys, the basic element that provides the security and privacy of our communications. The thesis starts from the following fundamental question: Does randomness exist in photonics? If so, how can we extract it and make it accessible to everyone? To address these two questions, tools from fundamental physics to engineering have been combined. The thesis starts from a detailed study of the phase diffusion process in semiconductor lasers and how to apply this process to the generation of random numbers. Unlike other physical processes based on deterministic laws of nature, phase diffusion has a purely quantum origin, and therefore, is an ideal source for generating random numbers. First, and using this phase diffusion process, we created the fastest quantum random number generator ever implemented (at that time) using only components from the telecommunications industry. More than 40 Gb/s were demonstrated using a pulsed laser scheme. Subsequently, we built several prototypes that were tested in fundamental science and supercomputing applications. In particular, some of the prototypes developed in this thesis were key in the famous loophole-free Bell tests experiments performed in 2015. In the process of building these prototypes, we started a new line of research to try to answer a new question: How do we know if the random numbers we generate really arise from the phase diffusion process? As a result, we introduced a new methodology, randomness metrology. Finally, we moved in the direction of miniaturization of the technology using techniques from the integrated photonics industry. In particular, we demonstrated the first fully integrated quantum random number generator, using a two-laser scheme on an Indium Phosphide chip. In parallel, we also demonstrated the integration of part of the device using Silicon technology, thus opening the door to large-scale production through the most advanced semiconductor industry.

In a society permeated by information and communication technologies, guaranteeing the security of data is of central importance. From exchanging messages with our friends to enabling businesses and governments share sensitive data, security is also of central importance in the development of the hyper-connected society. In this new technological paradigm, billions of Internet of Things devices will connect to the Internet, together with wearable devices and personalised health care systems. Self-driving cars will take us everywhere, and Industry 4.0 and Blockchain technologies will transform businesses and operations worldwide. However, without effective and long-term security guarantees, most of these new technologies will never reach the market. Current cyber-security technologies rely on unproven computational assumptions, compromising the privacy of our communications. The consequences of having all of our data exposed are unmeasurable, and there is an urgency to develop novel long-term solutions. Remarkably, quantum technologies offer unique opportunities for this endeavour. Quantum entropy sources and quantum key distribution systems can solve two of the most basic requirements in cryptography: the generation of unpredictable keys and their distribution. In this thesis, we have developed and demonstrated a highly scalable quantum entropy source with the potential of being integrated in any connected device, ultimately providing unprecedented security guarantees across any device. The same technology for securing communications can also be used in randomised algorithms, which are used to solve complex problems, such as in Monte Carlo methods for finance, science, or engineering, and in Machine Learning for Artificial Intelligence.

Randomness; Existence; Free Will; Evolution; Chance; Universe; Deterministic Machine; Information and Technology Era; Random Numbers; Communication Protocols; Ethernet; Classification Algorithms; Processing Algorithms; Page Rank; Monte Carlo Methods; Physics; Biology; Chemistry; Finance; Mathematics; Cryptography; Cybersecurity; Cryptographic Keys; Security; Privacy; Photonics; Phase Diffusion Process; Semiconductor Lasers; Quantum Origin; Quantum Random Number Generator; Telecommunications Industry; Pulsed Laser Scheme; Prototypes; Fundamental Science; Supercomputing Applications; Loophole-free Bell Tests; Randomness Metrology; Miniaturization; Integrated Photonics Industry; Fully Integrated Quantum Random Number Generator; Two-Laser Scheme; Indium Phosphide Chip; Silicon Technology; Large-Scale Production; Semiconductor Industry.