Precise estimates for certain distances in ℝ d \mathbb{R}^{d}

Matteo Fiacchi, Nikolai Nikolov · 2024

Read Precise estimates for certain distances in ℝ d \mathbb{R}^{d} by Matteo Fiacchi, Nikolai Nikolov (2024) 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-2412-10012.txt.

Introduction

Matteo Fiacchi and Nikolai Nikolov present a mathematics paper from 2024 on how distance is measured inside curved domains. Their work deals with Finsler metrics, which assign a length to paths in a way that can vary with direction. The focus is on sharp bounds for the built-in distances of these metrics, with close care paid to what happens near the edge of a domain.

Inside, the authors treat three named cases in turn. One is the Kobayashi-Hilbert metric near strongly convex points, and one is the minimal metric near convex points. A third case covers a quasi hyperbolic metric on strongly convex domains. The paper closes with a result that pins down when that third metric has a set form, which gives later work a firm test to use.

Catalog record

CREATORMatteo Fiacchi, Nikolai Nikolov
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE54 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-10012

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Precise estimates for certain distances in ℝ d \mathbb{R}^{d} Matteo Fiacchi, Nikolai Nikolov Abstract . We provide sharp estimates for the intrinsic distances of Finsler metrics with precise boundary estimates. These metrics include the Kobayashi-Hilbert metric near strongly convex points, the minimal metric near convex and strongly minimally convex points, and the k k -quasi hyperbolic metric in k k -strongly convex domains. Finally, we prove a characterization result in convex geometry for the k k -quasi hyperbolic metric. 2020 Mathematics Subject Classification 53C23, 32F45, 53A10 1. Introduction In their famous paper [ 1 ] , Balogh and Bonk proved the Gromov hyperbolicity of the Kobayashi distance k_{D} in strongly pseudoconvex domains, establishing the following estimates for the Kobayashi distance: if D\subset\mathbb{C}^{d} is a strongly pseudoconvex domain, then there exists B>0 such that g_{D}(z,w)-B\leq k_{D}(z,w)\leq g_{D}(z,w)+B,\ \ \ \ z,w\in D (1.1) where g_{D} is a function derived from the Carnot-Carathéodory metric on the boundary \partial D . This estimate, while sufficient to prove Gromov hyperbolicity, does not provide useful information about the distance when the points are close to each other. The estimates ( 1.1 ) were recently improved in [ 10 ] in the case of strongly pseudoconvex domains with \mathcal{C}^{2,\alpha} -smooth boundary. A similar situation arises in the real case, where the first named author in [ 5 ] proved the Gromov hyperbolicity of the minimal distance \rho_{D} (analogous to the Kobayashi metric in the theory of minimal surfaces) in strongly minimally convex domains by showing estimates similar to those of Balogh and Bonk: if D\subset\mathbb{R}^{d} is strongly minimally convex domain, then there exists A>1 and B>0 such that A^{-1}d_{D}(x,y)-B\leq\rho_{D}(x,y)\leq Ad_{D}(x,y)+B,\ \ \ \ x,y\in D where d_{D}(x,y)=2\log\left(\frac{\|\pi(x)-\pi(y)\|+h(x)\vee h(y)}{\sqrt{h(x)h(y)}}\right), (1.2) \pi\colon D\to\partial D is such that \|\pi(x)-x\|=d_{Euc}(x,\partial D) and h(x):=\sqrt{d_{Euc}(x,\partial D)} . Also in this case, the estimates were sufficient to prove Gromov hyperbolicity. The main goal of this paper is to improve the previous estimate in a way that can be applied to various distances naturally defined in real domains. These distances are all defined through a Finsler metric (see Subsection 2.1 for a brief introduction), with behavior at the boundary that is asymptotic to 1/2\delta_{D} in the normal component and comparable to 1/\delta_{D}^{1/2} in the tangential component, where \delta_{D}(\cdot):=d_{Euc}(\cdot,\partial D) . Let us now delve into the details. Let D\subset\mathbb{R}^{d} be a domain and \xi\in\partial D be a \mathcal{C}^{2} -smooth boundary point, then there exists a neighborhood U of \xi such that \pi\colon U\to\partial D is a well defined \mathcal{C}^{2} -smooth function. Set for x,y\in D\cap U a_{D}(x,y):=\frac{\|\pi(x)-\pi(y)\|+h(x)\vee h(y)}{\sqrt{h(x)h(y)}}-1, and, for c>0 , consider the quasi-distance defined as d^{c}_{D}(x,y)=2\log(1+ca_{D}(x,y)). Note that d^{1}_{D} is exactly the function that appears in ( 1.2 ). Let x\in D and v\in\mathbb{R}^{d} . If x is sufficiently close to a \mathcal{C}^{2} -smooth boundary point, we can decompose the vector v into its normal component v_{N} and its tangential component v_{T} at the boundary point \pi(x) . The main result of the paper is the following. Theorem 1.1 . Let D\subset\mathbb{R}^{d} (d\geq 2) be a domain and \xi\in\partial D be a \mathcal{C}^{2} -smooth boundary point. Let F\colon D\times\mathbb{R}^{d}\to[0,+\infty) be a Finsler metric on D and assume there exist a neighborhood U of \xi and 0<c_{1}<C_{1} such that for all x\in D\cap U and v\in\mathbb{R}^{d} \max\left\{(1-\omega(\delta_{D}(x)))\frac{\|v_{N}\|}{2\delta_{D}(x)},c_{1}\frac{\|v_{T}\|}{\delta_{D}(x)^{1/2}}\right\}\leq F(x,v)\leq(1+\omega(\delta_{D}(x)))\frac{\|v_{N}\|}{2\delta_{D}(x)}+C_{1}\frac{\|v_{T}\|}{\delta_{D}(x)^{1/2}} (1.3) where \omega\colon[0,\varepsilon]\to\mathbb{R} is a measurable function with \int_{0}^{\varepsilon}\frac{\omega(u)}{u}du<+\infty . Let d be the intrinsic distance of F , then there exist a neighborhood V\subset\subset U of \xi and 0<c_{2}\leq 1\leq C_{2} such that for all x,y\in D\cap V d^{c_{2}}_{D}(x,y)\leq d(x,y)\leq d^{C_{2}}_{D}(x,y). (1.4) The estimates of the main theorem are in the spirit of [ 10 ] , and they are effective regardless of the relative positions of the two points. In the second part of the paper, we will prove the estimates ( 1.3 ) for various Finsler metrics (1) Kobayashi-Hilbert metric near strongly convex points (Subsection 4.1 ); (2) minimal metric near convex and strongly minimally convex points (Subsection 4.2 ); (3) k -quasi hyperbolic metric in k -strongly convex domains (Subsection 4.3 ). As a consequence, we obtain the estimates ( 1.4 ) for the associated intrinsic distances. The paper concludes with a rigidity theorem for k -quasi-hyperbolic metrics (Theorem 5.1 ). Acknowledgements. The first author was partially supported by the European Union (ERC Advanced grant HPDR, 101053085 to Franc Forstnerič) and the research program P1-0291 from ARIS, Republic of Slovenia. The second author was partially supported by the Bulgarian National Science Fund, Ministry of Education and Science of Bulgaria under contract KP-06-N52/3. The first author acknowledges the Institute of Mathematics and Informatics for the hospitality during the visit in November 2024, during which this work was prepared. 2. Preliminaries Notations: • For x\in\mathbb{R}^{d} let ||x|| denote the standard Euclidean norm of x . • For u,v\in\mathbb{R}^{d} let \langle u,v\rangle denote the Euclidean scalar product of \mathbb{R}^{d} . • Let \mathbb{D}:=\{x\in\mathbb{R}^{2}:||x||<1\} be the unit disk in \mathbb{R}^{2} . • For x,y\in\mathbb{R}^{d} let (x,y):=\{tx+(1-t)y:t\in(0,1)\} be the open segment between x and y , and [x,y]:=\{tx+(1-t)y:t\in[0,1]\} be the closed segment between x and y . • For x\in\mathbb{R}^{d} and r>0 we denote with B(x,r):=\{y\in\mathbb{R}^{d}:\|x-y\|<r\} the Euclidean ball with center x and radius r . • For A,B\subseteq\mathbb{R}^{d} nonempty let denote d_{Euc}(A,B):=\inf\{\|x-y\|:x\in A,y\in B\}, the Euclidean distance between A and B . If A=\{x\} is a singleton, we simply write d_{Euc}(x,\cdot):=d_{Euc}(\{x\},\cdot) . • If D\subsetneq\mathbb{R}^{d} is a domain and x\in\mathbb{R}^{d} let \delta_{D}(x):=d_{Euc}(x,\partial D), be the distance to the boundary. • Let a,b\in\mathbb{R} , let denote a\wedge b:=\min\{a,b\} and a\vee b:=\max\{a,b\} . 2.1. Finsler metric In this section, we recall the definition of the Finsler metric. Let D\subset\mathbb{R}^{d} be a domain, a Finsler metric is a function F\colon D\times\mathbb{R}^{d}\simeq TD\to[0,+\infty) with the following properties (1) F is upper semicontinuous on D\times\mathbb{R}^{d} ; (2) For all x\in D , v\in\mathbb{R}^{d} and t\in\mathbb{R} F(x,tv)=|t|F(x,v). Given a Finsler metric F on D and a piecewise \mathcal{C}^{1} -smooth curve \gamma\colon[a,b]\to D , we can define the length of \gamma with respect to the metric F \ell_{F}(\gamma):=\int_{a}^{b}F(\gamma(t),\dot{\gamma}(t))dt. Finally, the intrinsic pseudodistance associated with F is defined as d(x,y)=\inf_{\gamma}\ell_{F}(\gamma),\ \ \ x,y\in D where the infimum is over all piecewise \mathcal{C}^{1} -smooth curve \gamma\colon[0,1]\to D with \gamma(0)=x and \gamma(1)=y . 2.2. \mathcal{C}^{2} -smooth boundary point In this section, we will review some results on the geometry of the domain near a \mathcal{C}^{2} -smooth boundary point. Let D\subsetneq\mathbb{R}^{d} be a domain and \xi\in\partial D be a boundary point. We way that \xi is a \mathcal{C}^{2} -smooth boundary point if there exists a \mathcal{C}^{2} -smooth local defining function \rho near \xi , i.e., there exists a neighborhood U of \xi and \rho\colon U\to\mathbb{R} a \mathcal{C}^{2} -smooth function, such that (1) D\cap U=\{x\in U:\rho(x)<0\} ; (2) \nabla\rho\neq 0 in \partial D\cap U=\{x\in U:\rho(x)=0\}. The vector n_{\xi}:=\frac{\nabla\rho(\xi)}{\|\nabla\rho(\xi)\|} is called unit outer normal at \xi . It does not depend on the choice of the \mathcal{C}^{2} -smooth local defining function \rho . We now state the local version of a well-known lemma. Lemma 2.1 . Let D\subset\mathbb{R}^{d} (d\geq 2) be a domain and \xi_{0}\in\partial D be a \mathcal{C}^{2} -smooth boundary point. Then there exists r,\varepsilon>0 such that if we set U:=\{x\in\mathbb{R}^{d}:d_{Euc}(x,\partial D\cap B(\xi_{0},r))<\varepsilon\} we have (1) every point \xi\in\partial D\cap U is a \mathcal{C}^{2} -smooth boundary point;

Read the complete text