Paper 2022/1490

Efficient Gaussian sampling for RLWE-based cryptography through a fast Fourier transform

Marcio Barbado Junior, Escola Politecnica da Universidade de Sao Paulo
Abstract

Quantum computing threatens classical cryptography, leading to the search for stronger alternatives. The cryptographic approach based on lattices is considered as a viable option. Schemes with that approach use Gaussian sampling, a design which brings along two concerns: efficiency and information leakage. This work addresses those concerns in the RLWE formulation, for digital signatures. Efficiency mitigation uses the central limit theorem, and the Walsh–Hadamard transform, whereas the information leakage risk is reduced via isochronous implementation. Up to \( 2^{23} \) samples are queried, and the results are compared against those of a cumulative distribution table sampler. Statistical metrics show the suitability of the presented sampler in a number of contexts.

Note: Previously published in the proceedings of SBSeg 2022.

Metadata
Available format(s)
PDF
Category
Applications
Publication info
Published elsewhere. XXII Simpósio Brasileiro de Segurança da Informação e de Sistemas Computacionais (SBSeg 2022)
DOI
10.5753/sbseg.2022.224430
Keywords
RLWE Ring learning with errors Discrete Gaussian sampling Central limit theorem Fast Walsh–Hadamard transform
Contact author(s)
mbarbado @ usp br
History
2022-10-30: approved
2022-10-30: received
See all versions
Short URL
https://ia.cr/2022/1490
License
Creative Commons Attribution
CC BY

BibTeX

@misc{cryptoeprint:2022/1490,
      author = {Marcio Barbado Junior},
      title = {Efficient Gaussian sampling for {RLWE}-based cryptography through a fast Fourier transform},
      howpublished = {Cryptology {ePrint} Archive, Paper 2022/1490},
      year = {2022},
      doi = {10.5753/sbseg.2022.224430},
      url = {https://eprint.iacr.org/2022/1490}
}
Note: In order to protect the privacy of readers, eprint.iacr.org does not use cookies or embedded third party content.