Anzeige
Erschienen in:
2003 | OriginalPaper | Buchkapitel
Lattice Reduction by Random Sampling and Birthday Methods
verfasst von : Claus Peter Schnorr
Erschienen in: STACS 2003
Verlag: Springer Berlin Heidelberg
Enthalten in: Professional Book Archive
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
We present a novel practical algorithm that given a lattice basis b1, ..., bn finds in O(n2( k/6 )k/4) average time a shorter vector than b1 provided that b1 is ( k/6 )n/(2k) times longer than the length of the shortest, nonzero lattice vector. We assume that the given basis b1, ..., b n has an orthogonal basis that is typical for worst case lattice bases. The new reduction method samples short lattice vectors in high dimensional sublattices, it advances in sporadic big jumps. It decreases the approximation factor achievable in a given time by known methods to less than its fourth-th root. We further speed up the new method by the simple and the general birthday method.