Nonempty interior of pinned distance and tree sets
Tainara Borges, Benjamin Foster, Yumeng Ou, Eyvindur Palsson · 2025
Read Nonempty interior of pinned distance and tree sets by Tainara Borges, Benjamin Foster, Yumeng Ou, Eyvindur Palsson (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-15709.txt.
Introduction
Tainara Borges, Benjamin Foster, Yumeng Ou, and Eyvindur Palsson study distance sets pinned at a single point of a fractal set, in a math paper from 2025. For a compact set, the pinned set lists distances from one fixed point to all others. A classic result says a thick enough set in high dimensions has a pinned set with interior points.
The team obtains the first nontrivial threshold for this problem in the plane. They also improve the known threshold in three dimensions over the classic result. A novel induction argument extends the findings from distances to tree patterns, where several points and links must all appear.
Catalog record
| CREATOR | Tainara Borges, Benjamin Foster, Yumeng Ou, Eyvindur Palsson |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 78 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-15709
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Nonempty interior of pinned distance and tree sets Tainara Borges, Benjamin Foster, Yumeng Ou, Eyvindur Palsson Abstract . For a compact set E ⊂ ℝ d E\subset\mathbb{R}^{d} , d ≥ 2 d\geq 2 , consider the pinned distance set Δ y ( E ) = { | x − y | : x ∈ E } \Delta^{y}(E)=\{|x-y|:x\in E\} . Peres and Schlag showed that if the Hausdorff dimension of E E is bigger than d + 2 2 \frac{d+2}{2} with d ≥ 3 d\geq 3 , then there exists a point y ∈ E y\in E such that Δ y ( E ) \Delta^{y}(E) has nonempty interior. In this paper we obtain the first non-trivial threshold for this problem in the plane, improving on the Peres–Schlag threshold when d = 3 d=3 , and we extend the results to trees using a novel induction argument. 1. Introduction and Main results The Falconer distance problem asks how large the Hausdorff dimension of a compact set E\subset\mathbb{R}^{d} , d\geq 2 , needs to be to guarantee that its distance set \Delta(E)=\{|x-y|:x,y\in E\} has positive Lebesgue measure. Falconer [ 11 ] showed through a construction that the dimensional threshold \frac{d}{2} is the best one can hope for and proved that \textnormal{dim}_{\mathcal{H}}(E)>\frac{d}{2}+\frac{1}{2} is sufficient. Bridging this gap, namely solving Falconer’s conjecture that the threshold \frac{d}{2} is sufficient, is one of the major open problems at the interface of harmonic analysis and geometric measure theory. Despite much progress in recent years, this conjecture still remains open. The pinned Falconer distance problem similarly asks how large the Hausdorff dimension of a compact set E\subset\mathbb{R}^{d} , d\geq 2 , needs to be to guarantee that there exists a point y\in E such that the pinned distance set at y , \Delta^{y}(E)=\{|x-y|:x\in E\}, has positive Lebesgue measure. Any dimensional threshold for the pinned Falconer distance problem immediately implies that the same dimensional threshold works for the unpinned original problem. This pinned variant was first studied by Peres and Schlag [ 35 ] who obtained a sufficient threshold of \textnormal{dim}_{\mathcal{H}}(E)>\frac{d}{2}+\frac{1}{2} . The original unpinned Falconer distance problem saw improvements by Bourgain [ 4 ] , Wolff [ 40 ] and Erdoğan [ 10 ] that left the best threshold at \textnormal{dim}_{\mathcal{H}}(E)>\frac{d}{2}+\frac{1}{3} for a while. Liu [ 27 ] connected the main proof techniques for the pinned and unpinned variants and matched the threshold \textnormal{dim}_{\mathcal{H}}(E)>\frac{d}{2}+\frac{1}{3} for the pinned problem. Since Liu’s work, all the best thresholds for the pinned and unpinned variants have been the same. With multiple improvements recently [ 8 , 20 , 5 , 6 , 7 ] the best thresholds now stand at \frac{5}{4} in the plane and at \frac{d}{2}+\frac{1}{4}-\frac{1}{8d+4} when d\geq 3 . One of the motivations behind the Falconer distance problem is a classic result of Steinhaus [ 38 ] , which states that if a set E\subset\mathbb{R}^{d} , d\geq 1 , has positive Lebesgue measure, then the set E-E=\{x-y:x,y\in E\} contains a neighborhood of the origin, which immediately implies that the distance set \Delta(E) has positive Lebesgue measure. One can view the Falconer distance problem as reducing the assumptions needed on the set E to obtain the conclusion on the distance set. In fact, from the Steinhaus result one not only concludes that the distance set has positive Lebesgue measure but also has nonempty interior. In that spirit, Mattila and Sjölin [ 28 ] showed that a compact set E\subset\mathbb{R}^{d} , d\geq 2 , satisfying \textnormal{dim}_{\mathcal{H}}(E)>\frac{d}{2}+\frac{1}{2} has the property that its distance set \Delta(E) has nonempty interior. In the pinned setting Peres and Schlag [ 35 ] showed that if \textnormal{dim}_{\mathcal{H}}(E)>\frac{d}{2}+1 then there exists a point y\in E such that the pinned distance set \Delta^{y}(E) has nonempty interior. Despite all the recent progress on the Falconer distance problem and its pinned variant, neither of the results on nonempty interior has been improved since the original papers. Furthermore, observe that unlike when only asking for a positive Lebesgue measure, there is a gap here between the pinned and unpinned results. In the plane when d=2 one has \frac{d}{2}+1=2 so the result on nonempty interior for the pinned distance set in [ 35 ] does not provide any information in that case. This leads us to the first main result of this paper, which establishes the first non-trivial threshold for nonempty interior of the pinned distance set in the plane. Theorem 1 . Let E\subset\mathbb{R}^{2} be a compact set with \textnormal{dim}_{\mathcal{H}}(E)>\frac{7}{4} . Then there exists y\in E such that \Delta^{y}(E) contains an interval. Many further variants of the Falconer distance problem exist. There are dimensional variants that obtain a lower bound on \textnormal{dim}_{\mathcal{H}}(\Delta(E)) given \textnormal{dim}_{\mathcal{H}}(E) , especially in the critical case of \textnormal{dim}_{\mathcal{H}}(E)=\frac{d}{2} . See for example the recent works in [ 7 ] and the breakthrough result of Shmerkin and Wang [ 36 ] as well as the references therein. One can look at other notions of distance aside from the classical Euclidean one. Iosevich, Mourgoglou and Taylor [ 25 ] showed that the dimensional threshold \textnormal{dim}_{\mathcal{H}}(E)>\frac{d}{2}+\frac{1}{2} for nonempty interior obtained by Mattila and Sjölin extends to a wide range of other metrics. Particularly important to this paper is a result by Iosevich and Liu [ 24 ] in the pinned setting, which recovers the results of Peres and Schlag as well as improving on some dimensional variants and extending to a variety of metrics. Their Fourier analytic techniques rely on local smoothing inequalities and are very different than the projection type methods of Peres and Schlag. The framework of Iosevich and Liu relies on L^{2} methods and thus only takes in local smoothing estimates that fit that framework. In this paper we upgrade their framework to handle more general L^{p}\rightarrow L^{p} local smoothing estimates, which requires delicate arguments replacing simple L^{2} orthogonality. Our next main theorem shows how the threshold for nonempty interior of the pinned distance set changes in \mathbb{R}^{d} given different L^{p}\rightarrow L^{p} local smoothing estimates. The hypothesis in the theorem that we can control a norm of a \partial_{t} derivative will be used in conjuction with one-dimensional Sobolev embedding (in the t variable) to prove a continuity result that implies the conclusion on nonempty interior of the pinned distance set. Theorem 2 . Let d\geq 2 and 2<p<\infty fixed. Assume that there exists \eta>0 such that the following local smoothing estimate holds (1) \left(\int_{1}^{2}\int_{\mathbb{R}^{d}}|\partial^{\frac{1}{p}+\varepsilon}_{t}A_{t}(f)(x)|^{p}dxdt\right)^{1/p}\lesssim_{\varepsilon}\|f\|_{L^{p}_{-\eta+2\varepsilon}(\mathbb{R}^{d})} for every \varepsilon>0 sufficiently small. Then for every compact set E\subset\mathbb{R}^{d} with \textnormal{dim}_{\mathcal{H}}(E)>d-\eta , there exists y\in E such that \Delta^{y}(E) contains an interval. More generally, given two compact sets E,F in \mathbb{R}^{d} satisfying \frac{1}{p^{\prime}}\textnormal{dim}_{\mathcal{H}}(E)+\frac{1}{p}\textnormal{dim}_{\mathcal{H}}(F)>d-\eta, there exists a probability measure \mu_{F} in F such that \Delta^{y}(E) contains an interval for \mu_{F} -almost every y\in F . If the local smoothing conjecture (see Conjecture 11 ) were true in all dimensions, Theorem 2 would yield the dimensional threshold \frac{d}{2}+1-\frac{1}{2d} , which would improve on the threshold \frac{d}{2}+1 obtained by Peres and Schlag in all dimensions. Theorem 1 realizes this optimal threshold when d=2 . We however note that it is widely believed that the true threshold should be \frac{d}{2} , as in the original Falconer distance problem. To further show how robust our result is, we obtain an improvement on the Peres-Schlag dimensional threshold when