An Improved BKW Algorithm For LWE With Binary Uniform Errors

Abstract

The Learning with Errors (LWE) Problem has received much attention since its introduction and been widely used in cryptography. However the error sampled from Gaussian distribution affects efficiency of application based on LWE, the LWE variants with errors taken from uniform distribution came into being in 2013, which has an asymptotic complexity analysis.An improved BKW algorithm has been proposed to the LWE problem with binary uniform errors, and a study of the complexity of the algorithm is given in the paper. As a result, new bounds are provided for the concrete hardness of cryptography based on the variant, and parameter choices for quiet a few important algorithm.