Abstract

In distributed computing a group of participants in a protocol wish to compute some function of their inputs. Secure multiparty computation extends this idea by allowing the participants to compute this function in the presence of an adversary which controls some subset of them. Moreover the participants in the protocol can compute the function in a private manner: that is without learning anything other than their output. In particular the inputs of all other parties are hidden from them. Yao’s protocol is a protocol that allows secure two-party computation in the presence of adversaries who are “honest but curious”, that is they will follow the protocol but will try to glean as much information as possible from doing so. In this talk we will formally define what it means for a protocol to be secure in the presence of such adversaries and then describe how we can construct Yao’s protocol. We will finish by sketching a proof of its security.