Homomorphic encryption in computing systems and environments

摘要

A transformation function that satisfies at least linearity and convolution can be used to encrypt data. The transformation function can, for example, be a DFT with one or more evaluation points that can be kept secret for encryption. The transformation function can effectively serve as a transform map and can be used to achieve fully homographic encryption in a system where encrypted data can be manipulated by applying one or more operations and the resulting encrypted data can be decrypted by applying the inverse of the transformation function and/or transformation map. A transformation function that satisfies at least linearity and convolution can be used for various applications, including, for example, private/public key decryption schemes, a signature schemes, database query and search schemes, as well as various applications of homomorphic operations.

主权项

1. A method, implemented at least partly by an encrypting device, for encrypting data, wherein the method comprises:
obtaining a transformation function that satisfies at least linearity and convolution; and
applying the transformation function, at least to first data, to generate and store, in a non-transitory computer readable storage medium, at least encrypted first data for the first data, wherein the transformation function satisfies a cyclic convolution modulo polynomial including one or more evaluation points as its roots, using the one or more evaluation points as one or more secret keys for encrypting the first data; inserting at least one secret or sensitive value at least into one component of a data signal; and; blinding the secret or sensitive value by adding blinding data into at least one component of a data signal for storing in the non-transitory computer readable storage medium.