Paper 2023/504 Original: eprint.iacr.org/2023/504 ↗

Private Computation Based On Polynomial Operation

Abstract

Privacy computing is a collection of a series of technical systems that intersect and integrate many disciplines such as cryptography, statistics, artificial intelligence, and computer hardware. On the premise of not exposing the original data, it can realize the fusion, sharing, circulation and calculation of data and its value in a manageable, controllable and measurable way. In the case of ensuring that the data is not leaked, it can achieve the purpose of making the data available and invisible, as well as the conversion and release of data value. We propose a privacy computing algorithm based on polynomial operations based on the unsolvable problem of high-degree polynomial modulo n. Encryptors can encrypt their own private information to generate ciphertext that can be calculated. This ciphertext can support any integer calculation including addition, multiplication, power calculation, etc. Different ciphertexts generated by the same key are fully homomorphic, and addition and multiplication operations can be performed between these ciphertexts. All calculations in our encryption system are carried out with polynomials as the medium, so the calculation efficiency is guaranteed. Our ciphertext can be provided to a third party for processing without revealing the encryption party's key and secret information.