Gr\"obner Bases over Algebraic Number Fields

TL;DR

This paper introduces a new efficient probabilistic method for computing Gr"obner bases over algebraic number fields by leveraging modular techniques and a divide-and-conquer approach, significantly improving performance.

Contribution

The paper presents a novel modular and parallelizable algorithm for Gr"obner basis computation over algebraic number fields, extending existing methods with a probabilistic, divide-and-conquer strategy.

Findings

Algorithm outperforms existing methods in timing benchmarks.

Effective for a wide range of algebraic number field cases.

Parallelizable approach enhances computational efficiency.

Abstract

Although Buchberger's algorithm, in theory, allows us to compute Gr\"obner bases over any field, in practice, however, the computational efficiency depends on the arithmetic of the ground field. Consider a field $K = \mathbb{Q}(\alpha)$, a simple extension of $\mathbb{Q}$, where $\alpha$ is an algebraic number, and let $f \in \mathbb{Q}[t]$ be the minimal polynomial of $\alpha$. In this paper we present a new efficient method to compute Gr\"obner bases in polynomial rings over the algebraic number field $K$. Starting from the ideas of Noro [Noro, 2006], we proceed by joining $f$ to the ideal to be considered, adding $t$ as an extra variable. But instead of avoiding superfluous S-pair reductions by inverting algebraic numbers, we achieve the same goal by applying modular methods as in [Arnold, 2003; B\"ohm et al., 2015; Idrees et al., 2011], that is, by inferring information in characteristic zero from information in characteristic $p > 0$. For suitable primes $p$, the minimal polynomial $f$ is reducible over $\mathbb{F}_p$. This allows us to apply modular methods once again, on a second level, with respect to the factors of $f$. The algorithm thus resembles a divide and conquer strategy and is in particular easily parallelizable. At current state, the algorithm is probabilistic in the sense that, as for other modular Gr\"obner basis computations, an effective final verification test is only known for homogeneous ideals or for local monomial orderings. The presented timings show that for most examples, our algorithm, which has been implemented in SINGULAR, outperforms other known methods by far.