A NEW MULTI-PARTY PRIVATE SET INTERSECTION PROTOCOL BASED on OPRFs

Abstract

In many crucial real-world applications, parties must jointly perform some secure multi-party computation (MPC) while keeping their inputs hidden from other parties. Private Set Intersection (PSI), the specific area of Multi-Party Computation, let the parties learn the intersection of their private data sets without sharing their secret data with others. For instance, a smartphone user downloads a messaging application, naturally, he wants to discover who are the other contacts that are using the same application. The naive and insecure solution is to send all contacts to the server to discover them. However, the user does not want to share his contacts with the application for privacy issues. To handle this, in recent years, companies and organizations start to use PSI to enhance privacy and security with a little cost of communication and computation. In this paper, we introduce a novel method to compute Private Set Intersection with multi parties where there are at least three or more parties participating in the protocol. By employing the Zero-Secret Sharing scheme and Oblivious Pseudo-Random Functions (OPRFs), parties securely calculate the intersection with computational and communication complexities which are both linear in the number of parties.