Description du projet
Développement de systèmes de preuve efficaces pour vérifier l’exactitude des calculs
Les progrès récents de l’informatique, tels que l’informatique en nuage et les chaînes de blocs, font naître un besoin fondamental de vérifier l’exactitude des calculs. Le projet FASTPROOF, financé par l’UE, identifiera et remédiera aux principaux problèmes et goulets d’étranglement qui limitent actuellement l’évolutivité des preuves, tout en s’appuyant sur des hypothèses cryptographiques bien fondées. En particulier, FASTPROOF relèvera le défi majeur de réduire la complexité de la preuve de l’exactitude pour qu’elle soit linéaire dans le temps de calcul. Il développera et étudiera des méthodes sécurisées permettant de générer des preuves non interactives, qui présentent des avantages uniques. Enfin, il développera des preuves dans lesquelles la quantité de mémoire utilisée est proportionnelle à la mémoire requise par le calcul d’origine.
Objectif
A proof-system is a protocol that enables a powerful prover to convince a weaker verifier of the validity of a computational statement. Proof-systems have been central to the theory of computing since its inception. Some of the most important results, concepts and open problems in this field revolve around notions of efficient proof-systems.
In recent years, due to the surging need for large scale computation coupled with advances such as cloud computing and blockchains, these computational proof-systems, which originated in the theory literature, are now being implemented and deployed also in practice. However, their widespread deployment is impeded by key bottlenecks on the theory side. The goal of the FASTPROOF project is to identify, study and mainly to resolve these bottlenecks:
1. Minimizing the amount of back and forth interaction between the prover and the verifier, aiming at proofs that are non-interactive. Non-interactive proofs offer game-changing advantages: in contrast to their interactive counterparts, such proofs can simply be posted online and shared between different clients.
2. In current proof-systems the time that it takes for a prover to prove correctness of a computation is significantly longer than the time that it takes to simply perform the computation. The second major challenge is to reduce the complexity of proving correctness to be linear in the computation time.
3. Making the memory, or space complexity, needed to generate the proof be proportional to the space requirements of the computation, in contrast to most protocols in the literature in which the space required to prove correctness is proportional to the running time of the computation.
The goal of the FASTPROOF project is to resolve these key challenges while relying on well-founded cryptographic assumptions. While the focus of this project is theoretical, we believe that it stands to have an important impact also on the future development of practical proof-systems.
Programme(s)
- HORIZON.1.1 - European Research Council (ERC) Main Programme
Thème(s)
Régime de financement
HORIZON-AG - HORIZON Action Grant Budget-BasedInstitution d’accueil
32000 Haifa
Israël