Probabilistic behaviour of lattice reduction algorithms - GREYC amacc Access content directly
Book Sections Year : 2010

Probabilistic behaviour of lattice reduction algorithms

Abstract

The general behavior of lattice reduction algorithms is far from beingwell understood. Indeed, many experimental observations, regarding the executionof the algorithms and the geometry of their outputs, pose challenging questions,which remain unanswered and lead to natural conjectures yet to be settled. This surveydescribes complementary approaches which can be adopted for analyzing thesealgorithms, namely, dedicated modeling, probabilistic methods, and a dynamicalsystems approach.We explain how a mixed methodology has already proved fruitfulfor small dimensions p, corresponding to the variety of Euclidean algorithms(p D 1) and to the Gauss algorithm (p D 2). Such small dimensions constitute animportant step in the analysis of lattice reduction in any (high) dimension, since thecelebrated LLL algorithm, due to Lenstra, Lenstra, and Lov´asz, precisely involves asequence of Gauss reduction steps on sublattices of a large lattice.
Fichier principal
Vignette du fichier
CHS-VALLEE-2010-2.pdf (3.84 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-01083878 , version 1 (18-11-2014)

Identifiers

  • HAL Id : hal-01083878 , version 1

Cite

Brigitte Vallée, Antonio Vera. Probabilistic behaviour of lattice reduction algorithms. The LLL Algorithm, 2010, 978-3-642-02295-1. ⟨hal-01083878⟩
90 View
31 Download

Share

Gmail Facebook X LinkedIn More