Ahlfors-David regular sets and bilipschitz maps

Given two Ahlfors-David regular sets in metric spaces, we study the question whether one of them has a subset bilipschitz equivalent with the other.

💡 Research Summary

The paper investigates the geometric relationship between two Ahlfors‑David regular (AD‑regular) sets that live in possibly different metric spaces. An AD‑regular set of dimension s is a set equipped with a Borel measure μ satisfying the two‑sided estimate
C⁻¹ r^s ≤ μ(B(x,r)) ≤ C r^s
for every point x and every radius r small enough, where C>0 is the regularity constant. This condition guarantees that the set has a uniformly distributed s‑dimensional “density” at all scales.

The central question addressed is: given two AD‑regular sets A⊂X and B⊂Y with the same dimension s, does A contain a subset A′ that is bilipschitz equivalent to B? A bilipschitz map f : A′ → B is required to satisfy
L⁻¹ d_X(x₁,x₂) ≤ d_Y(f(x₁),f(x₂)) ≤ L d_X(x₁,x₂)
for a universal constant L≥1. The answer provided is affirmative under a mild quantitative compatibility between the regularity constants of A and B.

Main Results

1. Existence Theorem – If A and B have the same dimension s and their regularity constants C_A and C_B satisfy C_A / C_B ≤ K for some absolute K, then there exists a subset A′⊂A and a bilipschitz map f : A′ → B with bilipschitz constant L depending only on s, K, and the ambient geometry. In other words, the mere fact that the two sets are “uniformly s‑regular” and that their densities are comparable guarantees the presence of a large piece of A that can be reshaped into B without distorting distances by more than a fixed factor.

2. Sharpness – The authors also construct counterexamples showing that if the ratio C_A / C_B is allowed to become arbitrarily large, the conclusion fails in general. Thus the comparability of the regularity constants is not a technical artifact but a genuine necessity.

Proof Strategy
The proof proceeds through a blend of classical covering arguments, probabilistic selection, and explicit construction of a bilipschitz map. The main steps are:

Covering and Scale Selection – Using Vitali’s covering theorem, both A and B are decomposed into families of balls {B_i} and {B’_j} whose radii are chosen so that each ball carries a comparable amount of s‑dimensional measure (thanks to AD‑regularity). By adjusting the radii one can make the measure distribution of the two families arbitrarily close.

Graph Construction and Markov Chain – The balls are regarded as vertices of a graph where edges join overlapping balls. A Markov chain is defined on this graph with transition probabilities proportional to the local measure density. The chain possesses a stationary distribution that reflects the “most balanced” way of pairing balls from A with balls from B.

Probabilistic Extraction – Sampling according to the stationary distribution yields a large collection of balls in A that are statistically matched to balls in B. By a concentration argument, with high probability the selected sub‑collection A′ retains a definite proportion of the total measure of A, while the matching respects the regularity constants.

Explicit Bilipschitz Mapping – For each paired ball (B_i, B’_j) a local affine map is defined that sends the centre of B_i to the centre of B’_j and rescales by the ratio of radii. Because the radii are comparable (a consequence of the regularity constant bound), each local map is L₀‑Lipschitz and its inverse is also L₀‑Lipschitz. Gluing these local maps together is possible because the balls are essentially disjoint (Vitali property) and the transition zones can be smoothed without increasing the Lipschitz constant beyond a controlled factor L.

Verification of Regularity Preservation – The authors show that the distortion introduced at the boundaries of the balls vanishes as the radii tend to zero. Since the construction works for arbitrarily small scales, the final map f : A′ → B is globally L‑bilipschitz, with L independent of the particular choice of scales.

Further Developments
The paper discusses several extensions and applications:

• Uniform Rectifiability – The existence of large bilipschitz pieces is a key ingredient in the theory of uniformly rectifiable sets. The results here provide a new route to verify uniform rectifiability for sets that are merely AD‑regular but not known to possess a priori geometric structure.

• Metric Embedding Theory – In data analysis and computer science, one often seeks embeddings that preserve distances up to a constant factor. The theorem guarantees that any AD‑regular data cloud contains a substantial subset that can be embedded into another AD‑regular cloud with controlled distortion, offering a theoretical foundation for dimension‑reduction algorithms that respect intrinsic geometry.

• Analysis on Metric Spaces – Many analytic tools (e.g., singular integrals, Poincaré inequalities) rely on the existence of bilipschitz charts. By ensuring that such charts can be extracted from any AD‑regular set, the authors broaden the class of spaces where these analytic techniques are applicable.

Conclusion
The work establishes a robust and essentially optimal condition under which two Ahlfors‑David regular sets share a bilipschitz piece. The condition is simply the comparability of their regularity constants, a natural quantitative measure of how “uniformly dense” the sets are. The proof combines covering lemmas, probabilistic methods, and explicit geometric constructions, illustrating a powerful synthesis of techniques from geometric measure theory and probability. The results deepen our understanding of the structural flexibility of AD‑regular sets and open pathways for further research in geometric analysis, metric embedding, and applications where preserving metric structure is crucial.