Succinct publicly auditable MPC with universal setup

MPC, zk-SNARKs, zero knowledge, auditable MPC, marlin

Kanjalkar, Sanket Makarand ; Miller ; Andrew E

关键词: MPC, zk-SNARKs, zero knowledge, auditable MPC, marlin;   

Others  :  https://www.ideals.illinois.edu/bitstream/handle/2142/108065/KANJALKAR-THESIS-2020.pdf?sequence=1&isAllowed=y

美国|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

Inrecentyears,Multipartycomputationasaservice(MPCSaaS)isgainingpopularity as a promising approach for building privacy-preserving communication systems.Although there has been significant improvement in the efficiency and robustness of such protocols, in this thesis, we argue that these properties might not be enough for many MPC applications.Specifically, we need unconditional public auditability, which means that everyone can check that a given (secure) computation was performed correctly, even in the scenario where all the participants involved in the computation are corrupted.We also desire this auditability to be public, meaning any party that did not participate in the computation can audit it.We improve upon the previous state of the art adaptive Trinocchio by Veeningen[1] in three ways:1) Our auditor has constant pairing cost 2) Our proof sizes are reduced by a factor of three, and 3) we do not rely on a circuit-specific setup.In this work, we show the first construction of auditable robust MPC that supports fast verification, succinct proof size with one time universal, and updatable setup.Importantly,we provide auditability without significantly compromising the performance of the underlying Shamir secret shared MPC protocol, i.e, adding auditability only incurs a linear computation overhead and constant round communication overhead.We implement and evaluate our construction report various performance metrics