Asynchronous unconditionally secure computation: An efficiency improvement

Consider a set of n players who wish to compute the value of a commonly agreed multi-variate function on their local inputs, whilst keeping their local inputs as private as possible even when a non-trivial subset of players collude and behave maliciously. This is the problem of secure multiparty computation. In this work, we present an efficient protocol for asynchronous secure computation tolerating a computationally unbounded Byzantine adversary that corrupts up to t < n/4 players.