Polynomial Multiplication Systolic Array for Homomorphic Encryption in Secure Network Communications

Homomorphic encryption is a type of encryption that allows computation to be done directly on encrypted data, without the need to perform any decryption in the process. Brakerski/Fan-Vercauteren (BFV) is a homomorphic encryption scheme that uses Ring Learning with Error (RLWE) problem and encrypts data in ring polynomial form. The encryption and decryption in this scheme involve high degree polynomial multiplication. In this paper, we propose a systolic array design to accelerate polynomial multiplication using convolution method. In implementation result, we show result area for 2x2 and 4x4 polynomial multiplication systolic array, as a proof of concept we show a Verilog simulation for 15 degrees polynomial multiplication, and design space exploration with variation in bit-size and systolic array matrix size. From our exploration we found optimized design size that can be implemented in PYNQ-ZI board. The homomorphic encryption is useful for secure network communications and cloud analytics.