On triviality of uniform Diophantine exponents of lattices

Oleg N. German
DOI: https://doi.org/10.46298/cm.11137
2023-05-11
Abstract:In this paper we prove that uniform Diophantine exponents of lattices attain only trivial values.
Number Theory
What problem does this paper attempt to address?
The problem that this paper attempts to solve is about the value - taking problem of the uniform Diophantine exponents of lattices. Specifically, the author Oleg German proved that the uniform Diophantine exponents of lattices only take trivial values (i.e., 0 or infinity). The following is a detailed interpretation: ### Research Background In the theory of homogeneous linear Diophantine approximation, given \(n\) linearly independent linear forms \(L_1,\ldots,L_n\) and \(d\) real variables (\(n\leqslant d\)), a fundamental problem is to study how small the vector \((L_1(u),\ldots,L_n(u))\) can be when \(u\) runs through all non - zero integral points. There are multiple ways to measure the "size" of this vector, including: - Using some norm (for example, the supremum norm) - Using a weighted norm - Considering the product of the absolute values of each component of the vector No matter which function is chosen to measure the size of the vector, the concept of Diophantine exponent will be naturally introduced, which is the simplest quantitative characteristic to describe the approximation property. The cases of \(n < d\) and \(n = d\) need to be distinguished: - The case of \(n < d\) provides multiple Diophantine exponents. - The case of \(n = d\) provides the Diophantine exponents of lattices. ### Main Problems of the Paper The main purpose of the paper is to study the possible values of the uniform Diophantine exponents of lattices. Specifically, the author explored the following two problems: 1. **Can the uniform Diophantine exponents of lattices only take trivial values?** 2. **If so, what are these trivial values?** ### Main Results The main results of the paper are as follows: **Theorem 2.3**: Let \(\Lambda\) be a full - rank lattice in \(\mathbb{R}^d\). If \(\Lambda\) is similar to a sublattice of \(\mathbb{Z}^d\) under the action of the non - degenerate diagonal operator group, then \(\hat{\omega}(\Lambda)=\infty\); otherwise, \(\hat{\omega}(\Lambda) = 0\). This means that the uniform Diophantine exponents of lattices are either 0 or infinity, and there are no other possible values. ### Proof Methods The author used Davenport's lemma (Theorem 3.1) to prove the main results. Davenport's lemma is essentially an estimate of the \(k\)-th successive minimum of lattices, which provides a method for constructing specific parallelepipeds, thus proving that there are no non - zero lattice points inside these parallelepipeds. By detailed analysis and application of Minkowski's convex body theorem, the author proved that when the lattice \(\Lambda\) does not contain non - zero points in certain specific subspaces, its uniform Diophantine exponent must be 0. ### Conclusions This paper solved the problem that the uniform Diophantine exponents of lattices can only take trivial values, and clarified the specific contents of these trivial values. This result provides an important theoretical basis for understanding the lattice exponents in Diophantine approximation theory.