CryptoDB
Exact Lattice Sampling from Non-Gaussian Distributions
Authors: | |
---|---|
Download: | |
Abstract: | We propose a new framework for (trapdoor) sampling over lattices. Our framework can be instantiated in a number of ways. It allows for example to sample from uniform, affine and “product affine” distributions. Another salient point of our framework is that the output distributions of our samplers are perfectly indistinguishable from ideal ones, in contrast with classical samplers that are statistically indistinguishable. One caveat of our framework is that all our current instantiations entail a rather large standard deviation. |
Video from PKC 2021
BibTeX
@article{pkc-2021-30985, title={Exact Lattice Sampling from Non-Gaussian Distributions}, booktitle={Public-Key Cryptography - PKC 2021}, publisher={Springer}, doi={10.1007/978-3-030-75245-3_21}, author={Maxime Plançon and Thomas Prest}, year=2021 }