Doctoral thesis

Correlations and security : genuinely multipartite nonlocality and free-energy cryptography

  • 2022

PhD: Università della Svizzera italiana

English An important takeaway of Bell’s famous theorem is that the notion of information is intrinsically dependent on its physics. To expose this idea, I first analyze a colourful game, the Red-Green-Blue no-signalling game, which is a strikingly simple and instructive illustration of Bell nonlocality and of its underlying framework of Local Operation and Shared Randomness (LOSR). I then extend the setting to more parties in a network, and approach the natural question of how to define nonlocality that is genuinely tripartite. I use the inflation method to devise a thought experiment demonstrating in a device-independent way that the quantum states GHZ and W are genuinely tripartite nonlocal. I also show that this statement can be generalized to more parties in order to reach the conclusion that "there are correlations in Nature that are genuinely N-partite nonlocal, for any N." This claim has been recently verified by three independent experimental teams, up to N=4. Turning towards another fundamental theory of physics, I examine the cryptographic possibilities offered by the second law of thermodynamics. I find that unconditionally secure cryptography is, in principle, possible in a model where the only limited resource is free energy. More precisely, I build proof-of-principle protocols for secret-key establishment and multi-party computation that are secure against adversaries whose bound in free energy is exponentially larger than the amount required by the honest players. While impractical, the model probes the limits of reversible computing and gives rise to an "almost-no-cloning" theorem that is reminiscent but different from the quantum no-cloning theorem.
Collections
Language
  • English
Classification
Computer science and technology
License
License undefined
Open access status
green
Identifiers
Persistent URL
https://n2t.net/ark:/12658/srd1324962
Statistics

Document views: 70 File downloads:
  • 2022INF013.pdf: 145