Fiat-Shamir: from practice to theory

We give new instantiations of the Fiat-Shamir transform using explicit, efficiently computable hash functions. We improve over prior work by reducing the security of these protocols to qualitatively simpler and weaker computational hardness assumptions. As a consequence of our framework, we obtain the following concrete results. 1) There exists a succinct publicly verifiable non-interactive argument system for log-space uniform computations, under the assumption that any one of a broad class of fully homomorphic encryption (FHE) schemes has almost optimal security against polynomial-time adversaries. The class includes all FHE schemes in the literature that are based on the learning with errors (LWE) problem. 2) There exists a non-interactive zero-knowledge argument system for in the common reference string model, under either of the following two assumptions: (i) Almost optimal hardness of search-LWE against polynomial-time adversaries, or (ii) The existence of a circular-secure FHE scheme with a standard (polynomial time, negligible advantage) level of security. 3) The classic quadratic residuosity protocol of [Goldwasser, Micali, and Rackoff, SICOMP ’89] is not zero knowledge when repeated in parallel, under any of the hardness assumptions above.