Development of a Testbed for Fully Homomorphic Encryption Solutions

Abstract

Homomorphic encryption is a computing trend that promises to query a computer-based system without giving personal data. By means of these methods, a software is fed with encrypted data and the result of the computation is the encryption of the result obtained as if clear data were input to the same software. As this paradigm is attracting public and private funds, the number of the different approaches (supported by different software libraries) is growing over time. As theoretical principles underlying these libraries are different, functional and non-functional behaviours are different, too. The present work proposes to design and develop a testbed for these kinds of algorithms, making possible for a final user to build his/her own final functions and to test them against the variation of some parameters. A concrete prototype of the presented architecture is reported, based on the Pyfhel Python library and the underlying SEAL encryption library.