Slow polynomial mixing, dynamical Borel-Cantelli lemma and Hausdorff dimension of dynamical diophantine sets

Basic Information

• Paper ID: 2505.21464
• Title: Slow polynomial mixing, dynamical Borel-Cantelli lemma and Hausdorff dimension of dynamical diophantine sets
• Author: E. Daviaud (University of Liège)
• Classification: math.NT (Number Theory), math.DS (Dynamical Systems), math.MG (Metric Geometry)
• Publication Date: October 11, 2025 (arXiv v2)
• Paper Link: https://arxiv.org/abs/2505.21464

Abstract

In this paper, the author establishes optimality results concerning the dynamical Borel-Cantelli lemma and the Hausdorff dimension of certain dynamical Diophantine sets.

Research Background and Motivation

Problem Background

1. Diophantine Approximation Theory: This research originates from Dirichlet's work in number theory, addressing the rate at which real numbers can be approximated by rationals. Dirichlet proved that for any irrational number x, there exist infinitely many pairs of coprime integers (p,q) such that |x - p/q| ≤ 1/q².
2. Dynamical Analogue: In dynamical systems, analogous questions naturally arise concerning orbital approximations. Given a measurable map T: ℝᵈ → ℝᵈ and a T-ergodic measure μ, one studies the measure and Hausdorff dimension of the set {x : ||Tⁿ(x₀) - x||∞ ≤ 1/nᵈ i.o.}.
3. Limitations of Existing Results:
   • Galatolo proved in 5 that exponentially mixing systems satisfy the dynamical Borel-Cantelli lemma
   • Results were established in 3 under (γ,φ)-mixing conditions, but the optimality of these mixing conditions remained unknown

Research Motivation

This paper aims to prove that the mixing conditions proposed in 3 are optimal in a certain sense, both for the dynamical Borel-Cantelli lemma and for obtaining dimension estimates of classical dynamical Diophantine sets.

Core Contributions

1. Improved the Galatolo-Rousseau-Saussol example: Proved that this example is (9, n ↦ C/nˢ)-mixing, where 0 < s ≤ 1.
2. Established optimality results: Proved that for Lebesgue-almost every x₀,

   dimₕ{x : ||Tⁿ(x₀) - x||∞ ≤ 1/n^(1/3) i.o.} < 3
   

3. Proved optimality of mixing conditions: Showed that (γ, n ↦ C/n^(1+ε))-mixing suffices to guarantee the Borel-Cantelli lemma and dimension formula, while (γ, n ↦ C/n^(1-ε))-mixing does not.

Methodology in Detail

Problem Formulation

The paper investigates two fundamental problems in dynamical systems (T,μ):

1. Dynamical Borel-Cantelli lemma: Determining when μ({x : Tⁿ(x) ∈ B(x₀,rₙ) i.o.}) equals 0 or 1
2. Computing the Hausdorff dimension of dynamical Diophantine sets

Core Concepts

(γ,φ)-Mixing Property

For an ergodic system (T,μ), if for every ball A, B and every n ∈ ℕ,

μ(T⁻ⁿ(A) ∩ B) ≤ γμ(A)μ(B) + φ(n)μ(A)

then the system is called (γ,φ)-mixing.

Diophantine Linear Form

For α ∈ ℝ², its Diophantine linear form is defined as:

γₗ(α) = inf{γ ≥ 0 : ∃C > 0, dℤ(α·k) ≥ C/||k||∞^γ, ∀k ∈ ℤ²}

Model Architecture

Skew Product Map Construction

Consider the map Sₐ: 𝕋¹ × 𝕋² → 𝕋³, defined by:

Sₐ(x,t) = (2x, t + αχ[0,1/2[(x))

where α ∈ ℝ² is a vector with rationally independent coordinates.

Choice of Special α

According to Proposition 2.1, there exist α₁,α₂ ∈ ℝ such that:

• The partial quotients of their continued fraction expansions satisfy specific relations
• γₗ(α₁,α₂) = 16

Technical Innovations

1. Refined mixing analysis: Through dyadic cube decomposition and Abel summation formulas, precise estimates of mixing rates are obtained.
2. Dimension calculation techniques: Utilizing continued fraction theory and Diophantine approximation, upper bounds for Hausdorff dimension are derived.
3. Optimality proof strategy: By constructing specific counterexamples, the sharpness of theoretical results is demonstrated.

Experimental Setup

Theoretical Verification

This is primarily a theoretical work, with results verified through rigorous mathematical proofs rather than numerical experiments.

Key Lemma Verification

• Theorem 3.1: Utilizes discrepancy estimates for rotations on the 2-dimensional torus
• Lemma 3.2: Establishes equivalence between different formulations of mixing conditions

Experimental Results

Main Results

Proof of Theorem 2.4

1. First part: Proves that (Sₐ,ℒ³) is (9, n ↦ C/nˢ)-mixing, where s = 1/γ, γ > 16.
2. Second part: For each x₀,

   dimₕ{x : ||Tⁿ(x₀) - x||∞ ≤ 1/n^(1/3) i.o.} < 3
   

Specific Estimates

Through Abel summation and dyadic decomposition:

ℒ³(S⁻ⁿₐ(A) ∩ B) ≤ 9ℒ³(A) × ℒ³(B) + η/n^(1/γ) × ℒ³(A)

Hausdorff Dimension Upper Bounds

Proposition 4.1 proves:

dimₕ{y ∈ 𝕋² : ||kα - y||∞ ≤ 1/k^(1/3) i.o.} < 2

Related Work

Main Research Directions

1. Classical Diophantine approximation: Jarnik-Besicovitch theorem and its generalizations
2. Dynamical Borel-Cantelli lemma: Work by Chernov-Kleinbock and Galatolo
3. Fractal geometry: Methods for computing Hausdorff dimension

Relationship to Related Work

• Improves upon the example in 10 by Galatolo-Saussol-Rousseau
• Completes the theoretical framework in 3
• Resonates with (γ,φ)-mixing theory in 9

Conclusions and Discussion

Main Conclusions

1. Proved the optimality of mixing conditions in 3
2. Established precise estimates for the Hausdorff dimension of dynamical Diophantine sets
3. Refined the dynamical Borel-Cantelli theory

Theoretical Significance

The results indicate that:

• (γ, n ↦ C/n^(1+ε))-mixing is the weakest sufficient condition for the Borel-Cantelli lemma
• The dimension formula dimₕ = 1/δ holds under optimal mixing conditions

Future Directions

1. Generalization to higher-dimensional settings
2. Investigation of other types of dynamical systems
3. Exploration of practical applications

In-Depth Evaluation

Strengths

1. Theoretical completeness: Provides a complete characterization of mixing condition optimality
2. Technical sophistication: Skillfully combines dynamical systems theory, Diophantine approximation, and fractal geometry
3. Sharp results: Demonstrates the exactness of theoretical boundaries
4. Methodological innovation: Improves existing constructions and provides new analytical techniques

Limitations

1. Construction specificity: Relies on specific skew product map constructions
2. Dimensional restrictions: Main results are limited to 3-dimensional tori
3. Technical complexity: Proofs involve multiple deep theories with high entry barriers

Impact

1. Theoretical contribution: Refines dynamical Diophantine theory
2. Methodological value: Techniques may be applicable to similar problems
3. Interdisciplinary bridge: Promotes integration of dynamical systems and number theory

Applicable Scenarios

The methods are suitable for:

1. Studying mixing properties in dynamical systems
2. Dynamical generalizations of Diophantine approximation
3. Computing dimensions of fractal sets

References

The paper cites 10 important references, including:

• Beresnevitch-Velani's mass transference principle
• Galatolo's work on the dynamical Borel-Cantelli lemma
• Koukoulopoulos-Maynard's results on the Duffin-Schaeffer conjecture
• Classical fractal geometry texts (Falconer)

This paper represents an important theoretical contribution to dynamical systems theory. Through ingenious constructions and deep analysis, it refines the dynamical Borel-Cantelli theory and provides new insights and methods for related fields.