scientific article; zbMATH DE number 1014500

From MaRDI portal
Publication:4338236





This survey article treats a number of problems leading to vanishing sums \(\sum^k_{i=1} a_i\zeta^{n_i} =0\) of roots of unity. To fix the notation, let \(\zeta\) be a root of unity of exact order \(Q\) and the \(a_i\) complex numbers. Suppose the relation is irreducible, that is no subsum vanishes, and normalise by taking \(n_k=0\) and \(a_k\neq 0\). If there is such a vanishing relation, then there is one with the \(a_i\) in \(\mathbb{Q} (\zeta)\), so suppose the \(a_i\) are algebraic and lie in a number field of degree \(d\). Then the author and Dvornicich showed that \(Q\) is bounded by means of the relation \[ \sum_{p|Q} \left({p-1 \over (p-1,d)} -1\right) \leq k-1. \] A sketch of the proof is given using results analogous to those arising in the theory of the addition of sequences. Using this analysis, it is possible to bound the maximum order of a torsion solution of \(f(\zeta_1, \zeta_2)=0\) in terms of the degree of the polynomial \(f\) and the degree of the field generated by its coefficients, with certain exceptional types of polynomials, namely \(aX^n+ bX^m\) and \(cX^nY^m +d\). More generally, it is possible to study the distribution of points of small height on subvarieties of \({\mathcal G}^n_m\). (The points of zero height are just the torsion points, whose coordinates are roots of unity.) Another generalisation leads to the so-called \(S\)-unit equations and, in some respects, the development mimics the grander scale of that theory.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4338236)