Loading...
Please wait, while we are loading the content...
Méthodes pour la réduction d’attaques actives à passives en cryptographie quantique
| Content Provider | Semantic Scholar |
|---|---|
| Author | Lamontagne, Philippe |
| Copyright Year | 2018 |
| Abstract | Quantum mechanics offers an undeniable advantage over classical mechanics for conceiving protocols for cryptographic tasks. However, it also enables complex attacks that make quantum protocols hard to analyse and cryptographers are thus always looking for new tools to simplify proofs. In this thesis, we introduce new techniques that simplify the analysis of quantum cryptographic protocols and we showcase their usefulness with some applications. Our techniques are used to show that, under certain conditions, the security of a protocol against an attacker that more or less follows the honest behaviour (passive) is a sufficient condition to show the security against arbitrary attacks (active). We first demonstrate a general technique for reducing the security of a quantum protocol against an attacker with some amount of quantum memory to its security against an attacker with no quantum memory at all. This technique is especially useful due to the fact that attacks that have access to a quantum memory are notoriously difficult to analyse. We then introduce a second technique for certifying that a quantum population is close to a given mixed reference state : given n quantum registers, it is possible to make sure that the state of these n registers is close to φ⊗n for some mixed state φ. This technique offers a general framework for analysing a common situation in quantum cryptography where we want to certify that a state prepared by a distrusted source is close to the honest state. Both our techniques allowed us to contribute to the advancement of knowledge in quantum cryptography. Our first result is the main tool used to solve two open problems : the first concerning the cryptographic power of the one bit cut-and-choose primitive, and the second concerning the security of the BCJL bit commitment protocol. Our second technique is used to prove the security of a protocol that implements the task of producing a common random bit string that is almost uniformly distributed, even if one of the parties is dishonest. Our protocol outperforms any classical protocol for this task. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://www-etud.iro.umontreal.ca/~lamontph/papiers/Lamontagne_Philippe_2017_these.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |