Facebook Google Plus Twitter YouTube Mailing List RSS Feed

Random numbers from quantum processes

In everyday life we tend to think of randomness and unpredictability as undesirable, something to fight against. But randomness is an important resource: something we need to be able to guarantee in applications ranging from the security of communication to the reliability of computations. Unfortunately randomness is not cheap: randomness generation is typically resource-intensive and hard to verify. Indeed, the science of randomness extraction occupies a large field of classical information science.

From its beginnings, quantum physics has been known to entail a fundamental element of randomness. This fact was long seen as a nuisance – physicists pointed out that “we cannot know position and momentum simultaneously” and “we cannot fully control the outcome of a measurement”. Over the past two decades, however, the interdisciplinary work of quantum physicists and information scientists has given rise to the field of quantum information science. In this field, these non-classical features of nature are recognized as valuable resources. If one masters quantum technologies, randomness comes essentially for free.

Quantum randomness exhibits two crucial features:

  • It is intrinsic. In the classical domain, unpredictability is due to limited knowledge or resources. By contrast, quantum randomness is irreducible to ignorance: puzzling as it may sound, some physical events are intrinsically unpredictable.
  • It can be guaranteed to be private. Take a classical process, the outcome of which can safely be considered unpredictable given our resources. A list of numbers generated by such a process would be considered a good random list. However, once created, the list can be copied. A user holding a black-box device cannot ascertain whether the device is generating a new list or simply reading a copy of a previously generated list. This is clearly a bottleneck for secrecy. By contrast, quantum processes can be proven to generate private lists: it is guaranteed by the nature of the physical process that the device is generating the list and nobody can possibly be holding a copy.

With Principal Investigators in experimental physics, theoretical physics and theoretical computer science, CQT is a unique environment to study and produce randomness from quantum processes.

We are currently considering three interdisciplinary research directions:

  1. Guaranteed private randomness: we aim at the generation of certified randomness from loophole-free Bell tests, its characterization, and the study of its niche applications.
  2. Simpler and faster randomness: If one is willing to trust some features of the device, randomness of quantum origin can be generated using criteria and setups that are less demanding than a loophole-free Bell test. This may rapidly lead to a technological edge.
  3. Randomness from complex quantum processes: Whereas the two directions above are concerned with systems with few degrees of freedom, here the research aims at identifying signatures and effects of quantum randomness in complex system, in connection with thermodynamical quantities.