원문정보
초록
영어
Due to the various characteristics from the pseudo random number generator or many kinds of deterministic devices such as arithmetic processing units, new principles and test schemes should be proposed for assessment of true random number generator. In this contribution, a novel viewpoint on designing a Normal distribution sampler applicable for implementing a homomorphic encryption system based on Ring-LWE crypto scheme is proposed. We suggest a Gaussian normal distribution sampler described with HDL to create uniformly distributed pseudo random numbers which will be used for generating non-symmetric key matrices and error matrices using an open-source AES cryptographic module. The implemented sampler can be conducted with high-speed clock frequency with its succinct critical delay paths as well.
목차
1. Introduction
2. The Ring-Learning-With-Errors Cryptosystem
2.1 Learning with Errors
2.2 Ring-LWE
2.3 Security Presumption of Normal Sampling
3. Homomorphic Encryption Using Ring-LWE Encryption
4. Gaussians
5. Gaussian on Lattices
6. AES Block Cipher
7. Conclusion
Acknowledgments
References