arXiv subjects

Alexander Nabutovsky

Publications and source records attributed to Alexander Nabutovsky.

18 recordsLinked to original sources

arXiv2026

Geodesic nets on the Euclidean plane and closed geodesic nets on Euclidean surfaces

We prove that if $M$ is a closed Riemannian surface of diameter $d$ and area $v$ with sectional curvature in the $[-1,1]$ interval, then a closed geodesic net of length $l$ has at most $f(l,d,v)$ branch points, where $f(l,d,v)=(400\bar{l})^{(180\bar{l})^4}$ for $\bar{l}= \max\{l, \frac{\exp(d)}{\min\{1, \frac{v}{4}\}}\}$ This answers a question posed by S. Becker-Kahn. We also prove that for each geodesic net in the Euclidean plane with at most n unbalanced (boundary) vertices such that all its unbalanced vertices have degree 1, the number of balanced vertices of degree $\ge 3$ (=branch points) does not exceed $(25n)^{2n^2}$. This answers a question posed in [GM] and [NP].

math.DG↗

arXiv2024

Small separators, upper bounds for $l^\infty$-widths, and systolic geometry

We investigate the dependence on the dimension in the inequalities that relate the Euclidean volume of a closed submanifold $M^n\subset \mathbb{R}^N$ with its $l^\infty$-width $W^{l^\infty}_{n-1}(M^n)$ defined as the infimum over all continuous maps $\phi:M^n\longrightarrow K^{n-1}\subset\mathbb{R}^N$ of $sup_{x\in M^n}\Vert \phi(x)-x\Vert_{l^\infty}$. We prove that $W^{l^\infty}_{n-1}(M^n)\leq const\ \sqrt{n}\ vol(M^n)^{\frac{1}{n}}$, and if the codimension $N-n$ is equal to $1$, then $W^{l^\infty}_{n-1}(M^n)\leq \sqrt{3}\ vol(M^n)^{\frac{1}{n}}$. As a corollary, we prove that if $M^n\subset \mathbb{R}^N$ is {\it essential}, then there exists a non-contractible closed curve on $M^n$ contained in a cube in $\mathbb{R}^N$ with side length $const\ \sqrt{n}\ vol^{\frac{1}{n}}(M^n)$ with sides parallel to the coordinate axes. If the codimension is $1$, then the side length of the cube is $4\ vol^{\frac{1}{n}}(M^n)$. To prove these results we introduce a new approach to systolic geometry that can be described as a non-linear version of the classical Federer-Fleming argument, where we push out from a specially constructed non-linear $(N-n)$-dimensional complex in $\mathbb{R}^N$ that does not intersect $M^n$. To construct these complexes we first prove a version of kinematic formula where one averages over isometries of $l^N_\infty$ (Theorem 3.5), and introduce high-codimension analogs of optimal foams recently discovered in [KORW] and [AK].

math.DG↗

arXiv2023

Boxing inequalities in Banach spaces and Riemannian manifolds

We prove the following result: For each closed $n$-dimensional manifold $M$ in a (finite or infinite-dimensional) Banach space $B$, and each positive real $m\leq n$ there exists a pseudomanifold $W^{n+1}\subset B$ such that $\partial W^{n+1}=M^n$ and ${\rm HC}_m(W^{n+1})\leq c(m){\rm HC}_m(M^n)$. Here ${\rm HC}_m(X)$ denotes the $m$-dimensional Hausdorff content, i.e the infimum of $\Sigma_i r_i^m$, where the infimum is taken over all coverings of $X$ by a finite collection of open metric balls, and $r_i$ denote the radii of these balls. In the classical case, when $B=\mathbb{R}^{n+1}$, this result implies that if $\Omega\subset R^{n+1}$ is a bounded domain, then for all $m\in (0,n]$ ${\rm HC}_m(\Omega)\leq c(m){\rm HC}_m(\partial \Omega)$. This inequality seems to be new despite being well-known and widely used in the case, when $m=n$ (Gustin's boxing inequality, [G]). The result is a corollary of the following more general theorem that strengthens a theorem in [LLNR]: For each compact subset $X$ in a Banach space $B$ and positive real number $m$ such that ${\rm HC}_m(X)\not= 0$ there exists a finite $(\lceil m\rceil-1)$-dimensional simplicial complex $K\subset B$, a continuous map $\phi:X\longrightarrow K$, and a homotopy $H:X\times [0,1]\longrightarrow B$ between the inclusion of $X$ and $\phi$ (regarded as a map into $B$) such that: (1) For each $x\in X$ $\Vert x-\phi(x)\Vert_B\leq c_1(m){\rm HC}_m^{\frac{1}{m}}(X)$; (2) ${\rm HC}_m(H(X\times [0,1]))\leq c_2(m){\rm HC}_m(X)$. A similar theorem can also be proven in the case when $B$ is a metric space with a linear contractibility function and applies to all compact sets $X$ with a controllably small ${\rm HC}_m$ in Riemannian manifolds $M^n$ with the sectional curvature bounded below, the volume bounded below by a positive number, and the diameter bounded above.

math.MG↗

arXiv2022

Geodesic nets on non-compact Riemannian manifolds

A geodesic flower is a finite collection of geodesic loops based at the same point $p$ that satisfy the following balancing condition: The sum of all unit tangent vectors to all geodesic arcs meeting at $p$ is equal to the zero vector. In particular, a geodesic flower is a stationary geodesic net. We prove that in every complete non-compact manifold with locally convex ends there exists a non-trivial geodesic flower.

math.DG↗

arXiv2021

Filling metric spaces

We prove a new version of isoperimetric inequality: Given a positive real $m$, a Banach space $B$, a closed subset $Y$ of metric space $X$ and a continuous map $f:Y \rightarrow B$ with $f(Y)$ compact $$\inf_FHC_{m+1}(F(X))\leq c(m)HC_m(f(Y))^{\frac{m+1}{m}},$$ where $HC_m$ denotes the $m$-dimensional Hausdorff content, the infimum is taken over the set of all continuous maps $F:X\longrightarrow B$ such that $F(y)=f(y)$ for all $y\in Y$, and $c(m)$ depends only on $m$. Moreover, one can find $F$ with a nearly minimal $HC_{m+1}$ such that its image lies in the $C(m)HC_m(f(Y))^{1\over m}$-neighbourhood of $f(Y)$ with the exception of a subset with zero $(m+1)$-dimensional Hausdorff measure. The paper also contains a very general coarea inequality for Hausdorff content and its modifications. As an application we demonstrate an inequality conjectured by Larry Guth that relates the $m$-dimensional Hausdorff content of a compact metric space with its $(m-1)$-dimensional Urysohn width. We show that this result implies new systolic inequalities that both strengthen the classical Gromov's systolic inequality for essential Riemannian manifolds and extend this inequality to a wider class of non-simply connected manifolds.

math.DG↗

arXiv2020

Sweepouts of closed Riemannian manifolds

We show that for every closed Riemannian manifold there exists a continuous family of $1$-cycles (defined as finite collections of disjoint closed curves) parametrized by a sphere and sweeping out the whole manifold so that the lengths of all connected closed curves are bounded in terms of the volume (or the diameter) and the dimension $n$ of the manifold, when $n \geq 3$. An alternative form of this result involves a modification of Gromov's definition of waist of sweepouts, where the space of parameters can be any finite polyhedron (and not necessarily a pseudomanifold). We demonstrate that the so-defined polyhedral $1$-dimensional waist of a closed Riemannian manifold is equal to its filling radius up to at most a constant factor. We also establish upper bounds for the polyhedral $1$-waist of some homology classes in terms of the volume or the diameter of the ambient manifold. In addition, we provide generalizations of these results for sweepouts by polyhedra of higher dimension using the homological filling functions. Finally, we demonstrate that the filling radius and the hypersphericity of a closed Riemannian manifold can be arbitrarily far apart.

math.DG↗

arXiv2019

Complexity of Unknotting of Trivial 2-knots

We construct families of trivial $2$-knots $K_i$ in $\mathbb{R}^4$ such that the maximal complexity of $2$-knots in any isotopy connecting $K_i$ with the standard unknot grows faster than a tower of exponentials of any fixed height of the complexity of $K_i$. Here we can either construct $K_i$ as smooth embeddings and measure their complexity as the ropelength (a.k.a the crumpledness) or construct PL-knots $K_i$, consider isotopies through PL knots, and measure the complexity of a PL-knot as the minimal number of flat $2$-simplices in its triangulation. These results contrast with the situation of classical knots in $\mathbb{R}^3$, where every unknot can be untied through knots of complexity that is only polynomially higher than the complexity of the initial knot.

math.MG↗

arXiv2019

Linear bounds for constants in Gromov's systolic inequality and related results

Let $M^n$ be a closed Riemannian manifold. Larry Guth proved that there exists $c(n)$ with the following property: if for some $r>0$ the volume of each metric ball of radius $r$ is less than $({r\over c(n)})^n$, then there exists a continuous map from $M^n$ to a $(n-1)$-dimensional simplicial complex such that the inverse image of each point can be covered by a metric ball of radius $r$ in $M^n$. It was previously proven by Gromov that this result implies two by now famous Gromov's inequalities: $Fill Rad(M^n)\leq c(n)vol(M^n)^{1\over n}$ and, if $M^n$ is essential, then also $sys_1(M^n)\leq 6c(n)vol(M^n)^{1\over n}$ with the same constant $c(n)$. Here $sys_1(M^n)$ denotes the length of a shortest non-contractible closed curve in $M^n$. We prove that these results hold with $c(n)=({n!\over 2})^{1\over n}\leq {n\over 2}$. We demonstrate that for essential Riemannian manifolds $sys_1(M^n) \leq n\ vol^{1\over n}(M^n)$. All previously known upper bounds for $c(n)$ were exponential in $n$. Moreover, we present a qualitative improvement: In Guth's theorem the assumption that the volume of every metric ball of radius $r$ is less than $({r\over c(n)})^n$ can be replaced by a weaker assumption that for every point $x\in M^n$ there exists a positive $\rho(x)\leq r$ such that the volume of the metric ball of radius $\rho(x)$ centered at $x$ is less than $({\rho(x)\over c(n)})^n$ (for $c(n)=({n!\over 2})^{1\over n}$). Also, if $X$ is a boundedly compact metric space such that for some $r>0$ and an integer $n\geq 1$ the $n$-dimensional Hausdorff content of each metric ball of radius $r$ in $X$ is less than $({r\over 4n})^n$, then there exists a continuous map from $X$ to a $(n-1)$-dimensional simplicial complex such that the inverse image of each point can be covered by a metric ball of radius $r$.

math.MG↗

arXiv2019

Geodesic Nets: Some Examples and Open Problems

Geodesic nets on Riemannian manifolds form a natural class of stationary objects generalizing geodesics. Yet almost nothing is known about their classification or general properties even when the ambient Riemannian manifold is the Euclidean plane or the round $2$-sphere. In the first half of this paper we survey some results and open questions (old and new) about geodesic nets on Riemannian manifolds. Many of these open questions are about geodesic nets on the Euclidean plane. The second half contains a partial answer for one of these questions, namely, a description of a new infinite family of geodesic nets on the Euclidean plane with 14 boundary (or unbalanced) vertices and arbitrarily many inner (or balanced) vertices of degree $\geq 3$.

math.MG↗

arXiv2017

Sizes of spaces of triangulations of 4-manifolds and balanced presentations of the trivial group

Let $M$ be any compact four-dimensional PL-manifold with or without boundary (e.g. the four-dimensional sphere or ball). Consider the space $T(M)$ of all simplicial isomorphism classes of triangulations of $M$ endowed with the metric defined as the minimal number of bistellar transformations required to transform one of two considered triangulations into the other. Our main result is the existence of an absolute constant $C>1$ such that for every $m$ and all sufficiently large $N$ there exist more than $C^N$ triangulations of $M$ with at most $N$ simplices such that pairwise distances between them are greater than $2^{2^{\ldots^{2^N}}}$ ($m$ times). This result follows from a similar result for the space of all balanced presentations of the trivial group. ("Balanced" means that the number of generators equals to the number of relations). This space is endowed with the metric defined as the minimal number of Tietze transformations between finite presentations. We prove a similar exponential lower bound for the number of balanced presentations of length $\leq N$ with four generators that are pairwise $2^{2^{\ldots^{2^N}}}$-far from each other. If one does not fix the number of generators, then we establish a super-exponential lower bound $N^{const\ N}$ for the number of balanced presentations of length $\leq N$ that are $2^{2^{\ldots^{2^N}}}$-far from each other.

math.GT↗

arXiv2015

Balanced presentations of the trivial group and four-dimensional geometry

We prove that 1) There exist infinitely many non-trivial codimension one "thick" knots in $\mathbb{R}^5$; 2) For each closed four-dimensional smooth manifold $M$ and for each sufficiently small positive $ε$ the set of isometry classes of Riemannian metrics with volume equal to $1$ and injectivity radius greater than $ε$ is disconnected; 3) For each closed four-dimensional $PL$-manifold $M$ and any $m$ there exist arbitrarily large values of $N$ such that some two triangulations of $M$ with $<N$ simplices cannot be connected by any sequence of $<M_m(N)$ bistellar transformations, where $M_m(N)=\exp(\exp(\ldots \exp (N)))$ ($m$ times).

math.MG↗

arXiv2014

Contracting the boundary of a Riemannian 2-disc

Let $D$ be a Riemannian 2-disc of area $A$, diameter $d$ and length of the boundary $L$. We prove that it is possible to contract the boundary of $D$ through curves of length $\leq L + 200d\max\{1,\ln {\sqrt{A}\over d} \}$. This answers a twenty-year old question of S. Frankel and M. Katz, a version of which was asked earlier by M.Gromov. We also prove that a Riemannian $2$-sphere $M$ of diameter $d$ and area $A$ can be swept out by loops based at any prescribed point $p\in M$ of length $\leq 200 d\max\{1,\ln{\sqrt{A}\over d} \}$. This estimate is optimal up to a constant factor. In addition, we provide much better (and nearly optimal) estimates for these problems in the case, when $A<<d^2$. Finally, we describe the applications of our estimates for study of lengths of various geodesics between a fixed pair of points on "thin" Riemannian $2$-spheres.

math.DG↗

arXiv2014

Lengths of three simple periodic geodesics on a Riemannian $2$-sphere

Let $M$ be a Riemannian $2$-sphere. A classical theorem of Lyusternik and Shnirelman asserts the existence of three distinct simple non-trivial periodic geodesics on $M$. In this paper we prove that there exist three simple periodic geodesics with lengths that do not exceed $20d$, where $d$ is the diameter of $M$. We also present an upper bound that depends only on the area and diameter for the lengths of the three simple periodic geodesics with positive indices that appear as minimax critical values in the classical proofs of the Lyusternik-Shnirelman theorem. Finally, we present better bounds for these three lengths for "thin" spheres, when the area $A$ is much less than $d^2$, where the bounds for the lengths of the first two simple periodic geodesics are asymptotically optimal, when ${A\over d^2}\longrightarrow 0$.

math.DG↗

arXiv2006

Betti numbers of finitely presented groups and very rapidly growing functions

Define the length of a finite presentation of a group $G$ as the sum of lengths of all relators plus the number of generators. How large can be the $k$th Betti number $b_k(G)=$ rank $H_k(G)$ providing that $G$ has length $\leq N$ and $b_k(G)$ is finite? We prove that for every $k\geq 3$ the maximum $b_k(N)$ of $k$th Betti numbers of all such groups is an extremely rapidly growing function of $N$. It grows faster that all functions previously encountered in Mathematics (outside of Logic) including non-computable functions (at least those that are known to us). More formally, $b_k$ grows as the third busy beaver function that measures the maximal productivity of Turing machines with $\leq N$ states that use the oracle for the halting problem of Turing machines using the oracle for the halting problem of usual Turing machines. We also describe the fastest possible growth of a sequence of finite Betti numbers of a finitely presented group. In particular, it cannot grow as fast as the third busy beaver function but can grow faster than the second busy beaver function that measures the maximal productivity of Turing machines using an oracle for the halting problem for usual Turing machines. We describe a natural problem about Betti numbers of finitely presented groups such that its answer is expressed by a function that grows as the fifth busy beaver function. Also, we outline a construction of a finitely presented group all of whose homology groups are either ${\bf Z}$ or trivial such that its Betti numbers form a random binary sequence.

math.GR↗

arXiv2006

Lengths of geodesics between two points on a Riemannian manifold

Let x and y be two (not necessarily distinct) points on a closed Riemannian manifold M of dimension n. According to a celebrated theorem by J.P. Serre there exist infinitely many geodesics between x and y. The length of the shortest of these geodesics is obviously less than the diameter of M. But what can be said about the length of the other geodesics? We conjecture that for every k there are k geodesics between x and y of length not exceeding kd, where d denotes the diameter of M.This conjecture is obviously true for round spheres and it is not difficult to prove it for all closed Riemannian manifolds with non-trivial torsion-free fundamental groups. In this paper we announce two further results in the direction of this conjecture. Our first result is that the length of the second shortest geodesic between x and y does not exceed 2nd. Our second result is that if n=2 and M is diffeomorphic to the two-dimensional sphere, then for every k every two points on M can be connected by k geodesics of length not exceeding $(k^2/2 + 3k/2 +2)d$.

math.DG↗

arXiv2002

Volume, diameter and the minimal mass of a stationary 1-cycle

In this paper we present upper bounds on the minimal mass of a non-trivial stationary 1-cycle. The results that we obtain are valid for all closed Riemannian manifolds. The first result is that the minimal mass of a stationary 1-cycle on a closed n-dimensional Riemannian manifold M^n is bounded from above by (n+2)!d/3, where d is the diameter of a manifold M^n. The second result is that the minimal mass of a stationary 1-cycle on a closed Riemannian manifold M^n is bounded from above by 2(n+2)!Fill Rad(M^n) and, as a corollary, by 2(n+2)!(n+1)n^n(n!)^{1/2}(vol(M^n))^{1/n}, where Fill Rad(M^n) is the filling radius of the manifold, and vol(M^n) is its volume.

math.DG↗

arXiv1997

Variational problems for Riemannian functionals and arithmetic groups

In this paper we introduce a new approach to variational problems on the space Riem(M^n) of Riemannian structures (i.e. isometry classes of Riemannan metrics) on any fixed compact manifold M^n of dimension n >= 5. This approach often enables one to replace the considered variational problem on Riem(M^n) (or on some subset of Riem(M^n)) by the same problem but on spaces Riem(N^n) for every manifold N^n from a class of compact manifolds of the same dimension and with the same homology as M^n but with the following two useful properties: (1) If νis any Riemannian structure on any manifold N^n from this class such that Ric_(N^n,ν) >= -(n-1), then the volume of (N^n,ν) is greater than one; and (2) Manifolds from this class do not admit Riemannian metrics of non-negative scalar curvature. As a first application we prove a theorem which can be informally explained as follows: Let M be any compact connected smooth manifold of dimension greater than four, M et(M) be the space of isometry classes of compact metric spaces homeomorphic to M endowed with the Gromov-Hausdorff topology, Riem_1(M) in M et(M ) be the space of Riemannian structures on M such that the absolute values of sectional curvature do not exceed one, and R_1(M) denote the closure of Riem_1(M) in M et(M ). Then diameter regarded as a functional on R_1(M) has infinitely many "very deep" local minima.

math.DG↗

arXiv1997

Algorithmic aspects of homeomorphism problems

We will describe some results regarding the algorithmic nature of homeomorphism problems for manifolds; in particular, the following theorem. Theorem 1: Every PL or smooth simply connected manifold M^n of dimension n at least 5 can be recognized among simply connected manifolds. That is, there is an algorithm to decide whether or not another simply connected manifold is Top, PL or Diff isomorphic to M. Moreover, an analogous statement is true for embeddings in codimension at least three: one can algorithmically recognize any given embedding of one simply connected manifold in another up to isomorphism of pairs, or up to isotopy, if the codimension of the embedding is not two.

math.GT↗