Secure matrix multiplication based on fully homomorphic encryption


연구 분야: Cryptography



학회: The Journal of Supercomputing


초록

Fully homomorphic encryption allows to perform arbitrary computation over encrypted data which has great practical implications in the secure outsourced computation on an untrusted computation environment. This paper investigates secure matrix multiplication based on fully homomorphic encryption. We propose an efficient secure matrix multiplication scheme for arbitrary matrix such as \(A_{m\times l}\times B_{l\times n}\) based on the hypercube structure. Our proposal is the first single-ciphertext and composable secure matrix multiplication scheme for arbitrary matrix based on fully homomorphic encryption. Our scheme takes only l homomorphic multiplications and experimental results show that it has excellent performance for the matrices of different dimensions.


Author Profile
Hai Huang

Information School Zhejiang Sci-Tech University Hangzhou 310018 Zhejiang China

China
Author Profile
Haoran Zong

Information School Zhejiang Sci-Tech University Hangzhou 310018 Zhejiang China

China

📄 논문 정보

발행 연도 2022년
인용수 0
출판 국가 China
사이트 Springer
좋아요 수 0

연관 논문 목록 (353건)