Practical Techniques for Secure 2-party Computation

Secure 2-party computation (2PC) allows two mutually distrustful parties to carry out a computation on their combined private data, and learn only the output of that computation. 2PC was first proposed in the 1980s by Yao, and was considered mostly a theoretical curiosity until the mid 2000s. Since then, techniques for 2PC have become faster by many orders of magnitude. In this lecture series, I will cover the landscape of practical 2PC and the specific techniques that have improved its efficiency. Regarding general-purpose 2PC, we will study garbled circuits, oblivious transfer extension, cut-and-choose. We will also study private set intersection as an exemplar of practical special-purpose 2PC.