Comparison of Protocols for Private Set Intersection and 2-Party Secure Computation

Abstract

This thesis presents a comparative study of various cryptographic protocols designed for 2-Party Secure Computation (2PSC) and, more precisely, for Private Set Intersection (PSI), with a particular focus on their implementation and performance under malicious adversary models. Among the protocols explored, the OPA-based and PSZ protocols were selected for detailed analysis due to their strong security guarantees and efficiency. Both protocols were implemented using Python and subjected to empirical testing. The OPA-based protocol employs algebraic constructions such as Oblivious Linear Function Evaluation (OLE), its enhanced version OLE+, and Oblivious Polynomial Addition (OPA), whereas the PSZ protocol leverages modern Oblivious Transfer (OT) extensions and Pseudo-Random Functions (PRFs). The thesis evaluates these protocols in terms of computational cost and communication overhead. The results offer a practical perspective on using such cryptographic primitives in real-world applications.

This thesis presents a comparative study of various cryptographic protocols designed for 2-Party Secure Computation (2PSC) and, more precisely, for Private Set Intersection (PSI), with a particular focus on their implementation and performance under malicious adversary models. Among the protocols explored, the OPA-based and PSZ protocols were selected for detailed analysis due to their strong security guarantees and efficiency. Both protocols were implemented using Python and subjected to empirical testing. The OPA-based protocol employs algebraic constructions such as Oblivious Linear Function Evaluation (OLE), its enhanced version OLE+, and Oblivious Polynomial Addition (OPA), whereas the PSZ protocol leverages modern Oblivious Transfer (OT) extensions and Pseudo-Random Functions (PRFs). The thesis evaluates these protocols in terms of computational cost and communication overhead. The results offer a practical perspective on using such cryptographic primitives in real-world applications.