Fully homomorphic encryption hardware accelerator based on systolic array and design method thereof

Abstract

The invention discloses a fully homomorphic encryption hardware accelerator based on a systolic array and a design method of the fully homomorphic encryption hardware accelerator. The hardware accelerator comprises a control module, a systolic array module, a bit expansion module, an addition module, a multiplication module and a tiling module. The control module controls the input of data and the operation of related modules; the systolic array module is responsible for carrying out multiply-accumulate operation on input R and A matrixes; the bit expansion module expands the decimal number into a binary number; the multiplication module is responsible for carrying out multiplication operation on the parameter u and the unit matrix IN; the addition module is responsible for performing addition operation; and the tiling module flattens not all {0, 1} elements of the ciphertext into {0, 1} again. The invention aims to improve the calculation efficiency, reduce the encryption time and better meet the real-time requirement, and meanwhile, the flow line output operation is adopted, so that the hardware resource utilization rate is further improved.

Description

technical field [0001] The invention belongs to the field of integrated circuit encryption, in particular to a fully homomorphic encryption hardware accelerator applied to a systolic array and a design method thereof. Background technique [0002] In the past few years, an important development of cryptography is the introduction of fully homomorphic encryption technology. In traditional encryption algorithms, including symmetric IDEA algorithm, asymmetric RSA algorithm and irreversible AES encryption algorithm, these encryption algorithms focus on data security. Storage security. With the development of the concept of cloud computing, more and more users transfer their personal data to the cloud and complete a series of operations with the help of the cloud. However, this brings about data privacy issues. Some clouds are not safe. Users There is a risk of the data being stolen, and at this time fully homomorphic encryption came into being. [0003] In 2009, Gentry first pr...

AI Technical Summary

Problems solved by technology

[0004] However, with the improvement of the security level, the size of the public key of the private key is also increasing, and the parallelism of the operation in the past is relatively low, resulting in a significant increase in the encryption time. When the security level is 20, the encryption time is about It takes 1.717s, obviously unable to meet the real-time requirements

Images