완전 동형 암호에서의 정밀한 맥스 풀링 연산
Precise Max-Pooling on Fully Homomorphic Encryption
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
완전동형암호는 암호화된 데이터에 대한 대수적 연산을 지원하며, 최근에는 최대값 함수 등의 비대수적 연산도 근사하는 방법이 연구되고 있다. 그러나 아직 4개 이상의 숫자에 대한 정밀한 맥스 풀링 근사 연구는 이루어지지 않았다. 본 연구에서는 최대값 함수 근사 다항식의 합성을 활용하여 정밀한 맥스 풀링 근사 기법을 제안하였으며, 이를 이론적으로 분석하여 높은 정밀도를 증명하였다. 실험 결과, 제안하는 근사 맥스 풀링은 1ms 이내의 작은 분할 실행 시간과 이론적 분석과 일치하는 높은 정밀도를 보여주었다.
Abstract
Fully homomorphic encryption enables algebraic operations on encrypted data, and recently, methods for approximating non-algebraic operations such as the maximum function have been studied. However, precise approximation of max-pooling operations for four or more numbers have not been researched yet. In this study, we propose a precise max-pooling approximation method using the composition of approximate polynomials of the maximum function and theoretically analyze its precision. Experimental results show that the proposed approximate max-pooling has a small amortized runtime of less than 1ms and high precision that matches the theoretical analysis.
참고문헌 (13)
- C. Gentry, "Fully homomorphic encryption using ideal lattices", Proceedings of the Forty-First Annual ACM Symposium on Theory of Computing, pp.169~178 (2009)
- J. Cheon, "Homomorphic encryption for arithmetic of approximate numbers", Proceedings of ASIACRYPT 2017, LNCS, 10624, pp.409~437 (2017)
- J. Cheon, "A full RNS variant of approximate homomorphic encryption", Proceedings of Selected Areas in Cryptography 2018, LNCS, 11349, pp.347~368 (2018)
- J. Cheon, "Numerical method for comparison on homomorphically encrypted numbers", Proceedings of ASIACRYPT 2019, LNCS, 11922, pp.415~445 (2019)
- J. Cheon, "Efficient homomorphic comparison methods with optimal complexity", Proceedings of ASIACRYPT 2020, LNCS, 12492, pp.221~256 (2020)
- E. Lee, "Minimax approximation of sign function by composite polynomial for homomorphic comparison", IEEE Transactions on Dependable and Secure Computing, 19(6), pp.3711~3727 (2021)
- R. Cilad-Bachrach, "Cryptonets : applying neural networks to encrypted data with high throughput and accuracy", Proceedings of International Conference on Machine Learning, 2016, pp.201~210 (2016)
- R. Dathathri, "CHET : an optimizing compiler for fully-homomorphic neural-network inferencing", Proceedings of the 40th ACM SIGPLAN Conference on Programming Language Design and Implementation, pp.142~156 (2019)
- "Lattigo v3" (2022)[URL]
- J. Lee, "Privacy-preserving machine learning with fully homomorphic encryption for deep neural network", IEEE Access, 10, pp.30039~30054 (2022)
- E. Lee, "Low-complexity deep convolutional neural networks on fully homomorphic encryption using multiplexed parallel convolutions", Proceedings of International Conference on Machine Learning 2022, pp.12403~12422 (2022)
- W. Jung, "Over 100x faster bootstrapping in fully homomorphic encryption through memory-centric optimization with GPUs", IACR Transactions on Cryptographic Hardware and Embedded Systems, 2021(4), pp.114~148 (2021)
- S. Kim, "BTS : an accelerator for bootstrappable fully homomorphic encryption", Proceedings of the 49th Annual International Symposium on Computer Architecture, pp.711~725 (2022)