| dc.description.abstract | Homomorphic encryption is a cryptographic technique which allows for computations on encrypted data. This type of cryptography helps respect the privacy of an entity that wishes to outsource computations on their data. There are different types of homomorphic encryption. The strongest type of homomorphic encryption is called Fully Homomorphic Encryption (FHE). FHE is considered the strongest type because it allows for arbitrary computations on encrypted data. In 2009 Craig Gentry proved that it was possible to build an FHE scheme. Since then several FHE schemes have been proposed. This includes the Brakerski/Fan-Vercauteren (BFV) scheme, which performs well compared to older schemes.
Being able to outsource computations with homomorphic encryption opens a world of wide-ranging applications. One of these applications is Private Set Intersection (PSI). PSI is the problem of finding the intersection between two parties’ sets without revealing non-intersecting elements. Like HE, PSI is also another area of research within cryptography with its own applications. PSI has for example been used to help with privacy for contact tracing and password leakage services. Several approaches have been utilized to develop PSI protocols. One of the more modern approaches is an FHE-based approach, which we have studied. As part of our work we implemented a PSI protocol utilizing the BFV FHE scheme. We also tested the implementation to gauge its performance. | |