Modified Schmidt games and Diophantine approximation with weights (Q846981)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Modified Schmidt games and Diophantine approximation with weights
scientific article

    Statements

    Modified Schmidt games and Diophantine approximation with weights (English)
    0 references
    0 references
    0 references
    16 February 2010
    0 references
    Let \(r_1, \dots, r_n >0\) with \(r_1+ \cdots + r_n = 1\). A vector \((x_1,\dots, x_n) \in {\mathbb R}^n\) is said to be an element of \(\text{Bad}(r_1, \dots, r_n)\) if there is a constant \(C > 0\) such that for any \((q_1, \dots, q_n) \in {\mathbb Z}^n \setminus \{0\}\), \[ \max_{1 \leq i \leq n} \left\{ \| q x_i \|^{1/r_i}\right\} \geq {C \over q}. \] When \(r_i = 1/n\) for all \(i\), this definition coincides with the usual definition of a badly approximable vector. The difference here is that the quality of approximation is allowed to vary in some specified way among the coordinates. In the paper under review, the authors prove that for any choice of \(r_1, \dots, r_n\) as above, the set \(\text{Bad}(r_1, \dots, r_n)\) is winning for a modification of the so-called Schmidt game. This implies in particular that the set is thick, \textit{i.e.}, its intersection with any open set in \({\mathbb R}^n\) has maximal Hausdorff dimension. It alsoimplies certain nice intersection properties and extends some of the results of \textit{S. Kristensen}, \textit{R. Thorn} and \textit{S. Velani} [Adv. Math. 203, No. 1, 132--169 (2006; Zbl 1098.11039)]. The methods applied in the present paper have their origin in the theory of dynamical systems, and depend on a modified version of the Schmidt game induced by a contracting automorphism on a connected Lie group. The paper ends in a nice section exploring the possible extensions and limitations of the method, as well as the connection with dynamical systems. Also, some remarks on a conjecture of Schmidt which states that \(\text{Bad}(1/3, 2/3) \cap \text{Bad}(2/3, 1/3) \neq \emptyset\) are given. The methods of the present paper fall short of a proof of this conjecture. However, a stronger version of this conjecture has since been proved by \textit{D. Badziahin}, \textit{A. D. Pollington} and \textit{S. L. Velani} [On a problem in simultaneous Diophantine approximation: Schmidt's conjecture, Ann. of Math. (2), to appear, see \url{http://annals.math.princeton.edu/articles/2165}].
    0 references
    0 references
    Diophantine approximation
    0 references
    badly approximable vectors
    0 references
    Schmidt's game
    0 references
    Hausdorff dimension
    0 references
    0 references
    0 references