Contextualizing Alternative Models of Secret Sharing

Abstract

A secret sharing scheme is a means of distributing information to a set of players such that any authorized subset of players can recover a secret and any unauthorized subset does not learn any information about the secret. In over forty years of research in secret sharing, there has been an emergence of new models and extended capabilities of secret sharing schemes. In this thesis, we study various models of secret sharing and present them in a consistent manner to provide context for each definition. We discuss extended capabilities of secret sharing schemes, including a comparison of methods for updating secrets via local computations on shares and an analysis of approaches to reproducing/repairing shares. We present an analysis of alternative adversarial settings which have been considered in the area of secret sharing. In this work, we present a formalization of a deniability property which is inherent to some classical secret sharing schemes. We provide new, game-based definitions for different notions of verifiability and robustness. By using consistent terminology and similar game-based definitions, we are able to demystify the subtle differences in each notion raised in the literature.