Cryptographic Computation: Secure Fault-Tolerant Protocols and the Public-Key Model

Abstract

We give a general procedure for designing correct, secure, and fault-tolerant cryptographic protocols for many parties, thus enlarging the domain of tasks that can be performed efficiently by cryptographic means. We model the most general sort of feasible adversarial behavior, and describe fault-recovery procedures that can tolerate it. Our constructions minimize the use of cryptographic resources. By applying the complexity-theoretic approach to knowledge, we are able to measure and control the computational knowledge released to the various users, as well as its temporal availability.