PhD defense Tristan Nemoz: Computational models in Quantum Cryptography
Monday 9 Nov., 2026, at 14:00 (Paris time) at Télécom Paris
Télécom Paris, 19 place Marguerite Perey F-91120 Palaiseau [getting there], amphi 5 and in videoconferencing
Jury
- Elham KASHEFI, Research Director, CNRS and Sorbonne Université (Examiner)
- Mario BERTA, Professor of Physics (W3) and Director, Institute for Quantum Information RWTH Aachen University (Reviewer)
- Marco Túlio QUINTINO, Associate professor, CNRS and Sorbonne Université (Reviewer)
- Alex BREDARIOL GRILO, Researcher, CNRS and Sorbonne Université (Examiner)
- Romain ALLÉAUME, Professor, Télécom Paris, France (Thesis Director)
- Peter BROWN, Associate Professor, Télécom Paris, France (Thesis Co-supervisor)
Abstract
A recent work of Dupuis devised a randomness generation procedure that should generate more secure randomness than any other existing procedure. However, the amount of randomness that one can extract depends on the so-called optimized sandwiched conditional Rényi entropy, which is notoriously difficult to compute. Our first main result is thus to compute the maximum value of this conditional entropy so as to evaluate the maximal amount of randomness produced by this procedure. In addition to this contribution, we identify an important open problem concerning this randomness generation model which we haven’t been able to solve.
Learn more
PRSs are ensembles of states that are simple enoughto be prepared efficiently, but still complex enough to be indistinguishable from random states to the eye of a computationally-bounded adversary. The security of a PRS construction is generally proved via that of approximate state designs. Loosely, these are ensemble of states which are indistinguishable from random states. We show a limit on the security of real-valued approximate state designs. To do so, we determine the real-valued approximate state design with the highest security. We conjecture that this limitation also applies to PRSs, and not only on the bounds obtained when proving their security using approximate state designs. Our results also have applications outside of cryptography, such as the estimation of real-valued states.
PRSs have been initially defined using the qudits framework, which is not well-suited for quantum communication with existing technology. We thus define PRCSs, our notion of PRSs adapted to the coherent states framework. Using this framework, we adapt a protocol from Coladangelo to the coherent states framework and provide a security proof in this framework. This protocol is the first experimentally feasible, key-exchange protocol that only assumes the existence of quantum-secure pseudorandom functions, while allowing to send multiple photons per round, unlike traditional QKD methods.
PRSs have been initially defined using the qudits framework, which is not well-suited for quantum communication with existing technology. We thus define PRCSs, our notion of PRSs adapted to the coherent states framework. Using this framework, we adapt a protocol from Coladangelo to the coherent states framework and provide a security proof in this framework. This protocol is the first experimentally feasible, key-exchange protocol that only assumes the existence of quantum-secure pseudorandom functions, while allowing to send multiple photons per round, unlike traditional QKD methods.