Almost sure convergence of cover times for ψ-mixing systems

Basic Information

• Paper ID: 2412.17425
• Title: Almost sure convergence of cover times for ψ-mixing systems
• Author: Boyuan Zhao
• Classification: math.DS (Dynamical Systems)
• Publication Date: December 2024 (arXiv preprint)
• Paper Link: https://arxiv.org/abs/2412.17425

Abstract

This paper investigates the cover time problem for topologically transitive systems on the unit interval, which measures the time required for an orbit to reach a certain resolution level in a repeller. The author introduces a novel dimensional concept—stretched Minkowski dimension—and proves that under mixing conditions, the asymptotic behavior of typical cover times is determined by the Minkowski dimension when finite, otherwise by the stretched Minkowski dimension. As an application, the author demonstrates that for countably infinite-branched affine maps, results using the standard Minkowski dimension fail to produce finite logarithmic limits for cover times, while the stretched version yields finite limits. Furthermore, due to the absence of mixing properties, cover times for irrational rotations are explicitly computed as a counterexample.

Research Background and Motivation

Problem Definition

The core problem studied in this paper concerns cover times in dynamical systems. Given a topologically transitive piecewise expanding Markov map system on the unit interval, for a point x in the repeller Λ, the cover time is defined as:

$\tau_r(x) := \inf\{k : \forall y \in \Λ, \exists j \leq k : d(f^j(x), y) < r\}$

This represents the minimum time required for the orbit {f^j(x)} to visit the r-neighborhood of all points in the repeller.

Research Significance

1. Theoretical Importance: Cover times are essential tools for understanding the ergodic properties of orbits in dynamical systems, closely related to the dimensional properties of the system
2. Applied Value: Important applications in chaotic games, iterated function systems, Brownian motion, and other fields
3. Dimension Theory: Establishes profound connections between cover times and geometric properties of measures (dimensions)

Limitations of Existing Research

1. Existing results primarily focus on expected cover times Eτ_r, with insufficient investigation of almost sure convergence
2. Traditional Minkowski dimension can be infinite in certain cases, failing to characterize the asymptotic behavior of cover times
3. Lack of in-depth analysis of non-mixing systems (e.g., irrational rotations)

Core Contributions

1. Introduction of Stretched Minkowski Dimension: Defines a novel dimensional concept to handle cases where traditional Minkowski dimension is infinite
2. Establishment of Almost Sure Convergence Theorem: Under ψ-mixing conditions, precisely characterizes the asymptotic behavior of cover times
3. Provision of Concrete Application Examples: Proves asymptotic properties of cover times for countably infinite-branched affine maps
4. Analysis of Non-mixing Cases: Demonstrates the necessity of mixing conditions through the example of irrational rotations
5. Extension to Flow Systems: Generalizes results to flow systems satisfying specific conditions

Methodology Details

Problem Setup

Study the asymptotic behavior of cover times for typical points in repellers under topologically transitive piecewise expanding Markov maps f: X → 0,1 and their invariant ergodic probability measures μ.

Core Concepts

Traditional Minkowski Dimension

For a measure μ, define: $M_μ(r) := \min_{x∈\text{supp}(μ)} μ(B(x,r))$

The upper and lower Minkowski dimensions are respectively: $\overline{\dim}_M(μ) := \limsup_{r→0} \frac{\log M_μ(r)}{\log r}, \quad \underline{\dim}_M(μ) := \liminf_{r→0} \frac{\log M_μ(r)}{\log r}$

Stretched Minkowski Dimension (Technical Innovation)

When traditional Minkowski dimension is infinite, define: $\overline{\dim}^s_M(μ) := \limsup_{r→0} \frac{\log|\log M_μ(r)|}{-\log r}, \quad \underline{\dim}^s_M(μ) := \liminf_{r→0} \frac{\log|\log M_μ(r)|}{-\log r}$

This definition possesses scalar multiplication invariance and characterizes the stretched exponential decay rate of M_μ(r).

Main Theorems

Theorem 1.1 (Finite Minkowski Dimension Case)

Let (f,μ) be a probability-preserving system where f is topologically transitive, Markov, and piecewise expanding. If $\overline{\dim}_M(μ) < ∞$, then for μ-almost every point x in the repeller:

$\limsup_{r→0} \frac{\log τ_r(x)}{-\log r} ≥ \overline{\dim}_M(μ), \quad \liminf_{r→0} \frac{\log τ_r(x)}{-\log r} ≥ \underline{\dim}_M(μ)$

If (f,μ) is exponentially ψ-mixing, then equality holds.

Theorem 1.4 (Infinite Minkowski Dimension Case)

If $\overline{\dim}_M(μ) = ∞$ but $0 < \underline{\dim}^s_M(μ), \overline{\dim}^s_M(μ) < ∞$, then:

1. For μ-almost every x∈Λ: $\liminf_{r→0} \frac{\log\log τ_r(x)}{-\log r} ≥ \underline{\dim}^s_M(μ), \quad \limsup_{r→0} \frac{\log\log τ_r(x)}{-\log r} ≥ \overline{\dim}^s_M(μ)$
2. If (f,μ) is exponentially ψ-mixing, then equality holds.

Technical Innovations

1. Extension of Dimensional Concepts: Stretched Minkowski dimension fills gaps in traditional dimension theory
2. Application of Mixing Conditions: Skillfully utilizes exponential ψ-mixing properties to establish precise asymptotic equalities
3. Symbolic Dynamics Methods: Transforms the problem into analysis on symbolic spaces through semi-conjugacy π: Σ → Λ

Experimental Setup

Theoretical Verification Examples

Example 3.1: Finite-Branch Gibbs-Markov Maps

For Gibbs-Markov maps with locally Hölder potential functions, since |Df| is uniformly bounded and the Gibbs measure is exponentially ψ-mixing, Theorem 1.1 applies to yield: $\lim_{r→0} \frac{\log τ_r(x)}{-\log r} = \dim_M(μ)$

Example 3.2: Infinite Full-Branch Affine Maps

Construct maps with parameter κ > 1: $f(x) = cn^κ(x - a_{n-1}), \quad x ∈ [a_{n-1}, a_n) = P_n$

where $c = ζ(κ) = \sum_{n∈ℕ} \frac{1}{n^κ}$.

Proposition 3.1: $\dim_M(μ) = ∞$, but $\dim^s_M(μ) = \frac{1}{κ-1}$.

This demonstrates that traditional dimension fails while stretched dimension yields meaningful results.

Experimental Results

Main Results

Precise Characterization in Finite Dimension Case

For Gibbs-Markov systems, exact asymptotic behavior of cover times is obtained: $\lim_{r→0} \frac{\log τ_r(x)}{-\log r} = \dim_M(μ)$

Breakthrough in Infinite Dimension Case

For countably infinite-branched affine maps, it is proven that: $\lim_{r→0} \frac{\log\log τ_r(x)}{-\log r} = \frac{1}{κ-1}$

This is the first finite asymptotic limit obtained in the case where traditional dimension is infinite.

Counterexample Analysis: Irrational Rotations

Theorem 5.4 (Cover Times for Irrational Rotations)

For irrational rotations of type η(θ) > 1: $\liminf_{r→0} \frac{\log τ_r(x)}{-\log r} = \dim_M(μ) = 1 < η = \limsup_{r→0} \frac{\log τ_r(x)}{-\log r}$

This result demonstrates:

1. Absence of mixing properties leads to inequality of upper and lower limits
2. Necessity of mixing conditions for obtaining precise asymptotic behavior

Extension to Flow Systems

Theorem 6.2

For probability-preserving flows {f_t} satisfying conditions (H1)-(H5): $\liminf_{r→0} \frac{\log τ_r(x)}{-\log r} ≥ \dim_M(ν) - 1$

If $\dim_M(ν) = \dim_M(μ) + 1$, then: $\limsup_{r→0} \frac{\log τ_r(x)}{-\log r} ≤ \dim_M(μ)$

Related Work

Historical Development

1. Matthews (1988): First study of expected cover times for Brownian motion
2. Bárány-Jurga-Kolossváry (2023): Cover times in chaotic games for iterated function systems
3. Jurga-Todd (2024): Cover times in one-dimensional dynamical systems

Uniqueness of This Paper's Contributions

1. First establishment of precise almost sure convergence results
2. Introduction of stretched dimension to handle infinite dimension cases
3. Systematic analysis of the role of mixing conditions

Conclusions and Discussion

Main Conclusions

1. Under ψ-mixing conditions, the asymptotic behavior of cover times is completely determined by the corresponding dimension
2. Stretched Minkowski dimension is an effective tool for handling cases where traditional dimension is infinite
3. Mixing properties are necessary for obtaining precise asymptotic behavior

Limitations

1. Results are primarily restricted to one-dimensional systems and specific flow systems
2. ψ-mixing conditions are relatively strong, limiting the scope of applications
3. Geometric interpretation of stretched dimension requires further investigation

Future Directions

1. Generalization to higher-dimensional systems and more general dynamical systems
2. Investigation of cover time behavior under weaker mixing conditions
3. Exploration of applications of stretched dimension to other problems

In-Depth Evaluation

Strengths

1. Theoretical Innovation: Introduction of stretched Minkowski dimension fills theoretical gaps
2. Technical Rigor: Proof methods combine measure theory, ergodic theory, and symbolic dynamics
3. Complete Results: Both positive results and counterexamples form a comprehensive theoretical framework
4. Applied Value: Provides new tools for understanding ergodic properties of complex dynamical systems

Weaknesses

1. Scope of Applicability: Main results are limited to one-dimensional piecewise expanding maps
2. Mixing Conditions: Exponential ψ-mixing conditions may be too strong for practical applications
3. Geometric Intuition: Geometric meaning of stretched dimension lacks sufficient intuitive explanation

Impact

1. Theoretical Contribution: Makes important contributions to dimension theory and dynamical systems theory
2. Methodological Value: The concept of stretched dimension may have applications in other fields
3. Practical Utility: Provides practical tools for analyzing coverage properties of complex systems

Applicable Scenarios

1. Analysis of ergodic properties of piecewise linear maps
2. Investigation of coverage problems in iterated function systems
3. Understanding geometric structures of chaotic systems

References

Main references include:

• Bárány, B., Jurga, N. and Kolossváry, I. (2023). On the Convergence Rate of the Chaos Game
• Jurga, N., Todd, M. (2024). Cover times in dynamical systems
• Matthews, P. (1988). Covering problems for Brownian motion on spheres
• Bowen, R. (1975). Equilibrium States and The Ergodic Theory of Anosov Diffeomorphisms

This paper makes important contributions to dynamical systems theory, particularly at the intersection of cover times and dimension theory. The introduction of stretched Minkowski dimension not only resolves limitations of traditional methods but also opens new directions for future research.