[Paper Review] LOCAL HOMOGENEITY AND DIMENSIONS OF MEASURES IN DOUBLING METRIC SPACES

K Antti|arXiv (Cornell University)|Mar 15, 2010

Mathematical Dynamics and Fractals50 references9 citations

TL;DR

This paper introduces local homogeneity and the local L^q-spectrum as tools to analyze the local structure of measures in doubling metric spaces, enabling sharper estimates for local dimensions, conical densities, and the dimension of porous measures, thereby advancing multifractal analysis in a highly general geometric setting.

ABSTRACT

AKI, TAPIO RAJALA, AND VILLE SUOMALA Abstract. We introduce two new concepts, local homogeneity and local L q - spectrum, both of which are tools that can be used in studying the local structure of measures. The main emphasis is given to the study of local dimensions of measures in doubling metric spaces. As an application, we reach a new level of generality and obtain new estimates for conical densities, in multifractal analysis, and on the dimension of porous measures.

Motivation & Objective

• To develop new tools—local homogeneity and local L^q-spectrum—for analyzing the local structure of measures in metric spaces.
• To extend existing results on local dimensions and conical densities to the general setting of doubling metric spaces.
• To provide improved dimension estimates for porous measures using the introduced concepts.
• To generalize multifractal analysis techniques beyond Euclidean spaces to more general metric measure spaces.
• To establish a framework that captures fine-scale geometric and measure-theoretic properties of measures locally.

Proposed method

• Define local homogeneity as a condition on how uniformly a measure distributes mass in small neighborhoods.
• Introduce the local L^q-spectrum as a refinement of the global L^q-spectrum, capturing local scaling behavior.
• Use the interplay between local homogeneity and the local L^q-spectrum to derive bounds on local Hausdorff dimensions.
• Apply the theory to doubling metric spaces by leveraging the geometric control provided by the doubling condition.
• Establish connections between local homogeneity and conical density properties to estimate dimension via covering arguments.
• Use the local L^q-spectrum to analyze the dimension of porous measures by relating porosity to decay rates of local measure distributions.

Experimental results

Research questions

• RQ1 How can local homogeneity be formalized to capture the uniformity of measure distribution in small scales?
• RQ2 In what way does the local L^q-spectrum refine the global L^q-spectrum in describing local measure behavior?
• RQ3 To what extent can conical density estimates be improved using local homogeneity in doubling metric spaces?
• RQ4 How do local dimension estimates for porous measures depend on the local scaling properties of the measure?
• RQ5 What is the relationship between local homogeneity and the existence of tangent measures in doubling spaces?

Key findings

• Local homogeneity provides a new criterion to characterize the regularity of measure distribution at small scales in doubling metric spaces.
• The local L^q-spectrum allows for a more precise description of local scaling behavior than the global L^q-spectrum.
• Improved conical density estimates are obtained by combining local homogeneity with the doubling condition.
• The dimension of porous measures in doubling metric spaces is bounded using the local L^q-spectrum and local homogeneity, yielding sharper estimates than previous methods.
• The framework unifies and generalizes existing results in multifractal analysis by extending them to non-Euclidean doubling metric spaces.
• The paper establishes that local homogeneity implies favorable decay properties in measure concentration, which are essential for dimension estimates.