arXiv subjects

Ofer Neiman

Publications and source records attributed to Ofer Neiman.

At least 19 recordsLinked to original sources

arXiv2026

Path-Reporting Distance Oracles for Vertex-Labeled Graphs

Let $G=(V,E)$ be a weighted undirected graph, with $n$ vertices. A distance oracle is a data structure that can quickly answer distance queries, with some stretch factor. A seminal work of \cite{TZ01}, given an integer $k\ge 1$, provides such an oracle with stretch $2k-1$, query time $O(k)$, and size $O(k\cdot n^{1+1/k})$. Furthermore, this oracle can also report a path in $G$ corresponding to the returned distance. In this paper we focus on vertex-labeled graphs, in which each vertex is given a label from a set $L$ of size $\ell$. A {\em vertex-label distance oracle} answers queries of the form $(v,\lambda)$, where $v\in V$ and $\lambda\in L$, by reporting (an approximation to) the distance from $v$ to the closest vertex of label $\lambda$. Following \cite{HLWY11}, it was shown in \cite{C12} that for any integer $k> 1$, there exists a vertex-label distance oracle with stretch $4k-5$, query time $O(k)$, and size $O(k\cdot n\cdot \ell^{1/k})$. This state-of-the-art result suffers from two main drawbacks: The stretch is roughly a factor of 2 larger than in \cite{TZ01}, and it is not path-reporting. We address these concerns in this work, and provide the following results: First, we devise a {\em path-reporting} vertex-label distance oracle, at the cost of a slight increase in stretch and size. For any constant $0<\epsilon<1$, our oracle has stretch $(4k-5)\cdot(1+\epsilon)$, query time $O(k)$, and size $O(n^{1+o(1)}\cdot \ell^{1/k})$. Second, we show how to improve the stretch to the optimal $2k-1$, at the cost of mildly increasing the query time. Specifically, we devise a vertex-label distance oracle with stretch $2k-1$, query time $O(\ell^{1/k}\cdot\log n)$, and size $O(k\cdot n\cdot \ell^{1/k})$. \end{itemize}

cs.DS↗

arXiv2024

Lightweight Near-Additive Spanners

An $(\alpha,\beta)$-spanner of a weighted graph $G=(V,E)$, is a subgraph $H$ such that for every $u,v\in V$, $d_G(u,v) \le d_H(u,v)\le\alpha\cdot d_G(u,v)+\beta$. The main parameters of interest for spanners are their size (number of edges) and their lightness (the ratio between the total weight of $H$ to the weight of a minimum spanning tree). In this paper we focus on near-additive spanners, where $\alpha=1+\varepsilon$ for arbitrarily small $\varepsilon>0$. We show the first construction of {\em light} spanners in this setting. Specifically, for any integer parameter $k\ge 1$, we obtain an $(1+\varepsilon,O(k/\varepsilon)^k\cdot W(\cdot,\cdot))$-spanner with lightness $\tilde{O}(n^{1/k})$ (where $W(\cdot,\cdot)$ indicates for every pair $u, v \in V$ the heaviest edge in some shortest path between $u,v$). In addition, we can also bound the number of edges in our spanner by $O(kn^{1+3/k})$.

cs.DS↗

arXiv2024

Path-Reporting Distance Oracles with Linear Size

Given an undirected weighted graph, an (approximate) distance oracle is a data structure that can (approximately) answer distance queries. A {\em Path-Reporting Distance Oracle}, or {\em PRDO}, is a distance oracle that must also return a path between the queried vertices. Given a graph on $n$ vertices and an integer parameter $k\ge 1$, Thorup and Zwick \cite{TZ01} showed a PRDO with stretch $2k-1$, size $O(k\cdot n^{1+1/k})$ and query time $O(k)$ (for the query time of PRDOs, we omit the time needed to report the path itself). Subsequent works \cite{MN06,C14,C15} improved the size to $O(n^{1+1/k})$ and the query time to $O(1)$. However, these improvements produce distance oracles which are not path-reporting. Several other works \cite{ENW16,EP15} focused on small size PRDO for general graphs, but all known results on distance oracles with linear size suffer from polynomial stretch, polynomial query time, or not being path-reporting. In this paper we devise the first linear size PRDO with poly-logarithmic stretch and low query time $O(\log\log n)$. More generally, for any integer $k\ge 1$, we obtain a PRDO with stretch at most $O(k^{4.82})$, size $O(n^{1+1/k})$, and query time $O(\log k)$. In addition, we can make the size of our PRDO as small as $n+o(n)$, at the cost of increasing the query time to poly-logarithmic. For unweighted graphs, we improve the stretch to $O(k^2)$. We also consider {\em pairwise PRDO}, which is a PRDO that is only required to answer queries from a given set of pairs ${\cal P}$. An exact PRDO of size $O(n+|{\cal P}|^2)$ and constant query time was provided in \cite{EP15}. In this work we dramatically improve the size, at the cost of slightly increasing the stretch. Specifically, given any $\epsilon>0$, we devise a pairwise PRDO with stretch $1+\epsilon$, constant query time, and near optimal size $n^{o(1)}\cdot (n+|{\cal P}|)$.

cs.DS↗

arXiv2023

On the Size Overhead of Pairwise Spanners

Given an undirected possibly weighted $n$-vertex graph $G=(V,E)$ and a set $\mathcal{P}\subseteq V^2$ of pairs, a subgraph $S=(V,E')$ is called a ${\cal P}$-pairwise $\alpha$-spanner of $G$, if for every pair $(u,v)\in\mathcal{P}$ we have $d_S(u,v)\leq\alpha\cdot d_G(u,v)$. The parameter $\alpha$ is called the stretch of the spanner, and its size overhead is define as $\frac{|E'|}{|{\cal P}|}$. A surprising connection was recently discussed between the additive stretch of $(1+\epsilon,\beta)$-spanners, to the hopbound of $(1+\epsilon,\beta)$-hopsets. A long sequence of works showed that if the spanner/hopset has size $\approx n^{1+1/k}$ for some parameter $k\ge 1$, then $\beta\approx\left(\frac1\epsilon\right)^{\log k}$. In this paper we establish a new connection to the size overhead of pairwise spanners. In particular, we show that if $|{\cal P}|\approx n^{1+1/k}$, then a ${\cal P}$-pairwise $(1+\epsilon)$-spanner must have size at least $\beta\cdot |{\cal P}|$ with $\beta\approx\left(\frac1\epsilon\right)^{\log k}$ (a near matching upper bound was recently shown in \cite{ES23}). We also extend the connection between pairwise spanners and hopsets to the large stretch regime, by showing nearly matching upper and lower bounds for ${\cal P}$-pairwise $\alpha$-spanners. In particular, we show that if $|{\cal P}|\approx n^{1+1/k}$, then the size overhead is $\beta\approx\frac k\alpha$. A source-wise spanner is a special type of pairwise spanner, for which ${\cal P}=A\times V$ for some $A\subseteq V$. A prioritized spanner is given also a ranking of the vertices $V=(v_1,\dots,v_n)$, and is required to provide improved stretch for pairs containing higher ranked vertices. By using a sequence of reductions, we improve on the state-of-the-art results for source-wise and prioritized spanners.

cs.DS↗

arXiv2023

Light, Reliable Spanners

A \emph{$\nu$-reliable spanner} of a metric space $(X,d)$, is a (dominating) graph $H$, such that for any possible failure set $B\subseteq X$, there is a set $B^+$ just slightly larger $|B^+|\le(1+\nu)\cdot|B|$, and all distances between pairs in $X\setminus B^+$ are (approximately) preserved in $H\setminus B$. Recently, there have been several works on sparse reliable spanners in various settings, but so far, the weight of such spanners has not been analyzed at all. In this work, we initiate the study of \emph{light} reliable spanners, whose weight is proportional to that of the Minimum Spanning Tree (MST) of $X$. We first observe that unlike sparsity, the lightness of any deterministic reliable spanner is huge, even for the metric of the simple path graph. Therefore, randomness must be used: an \emph{oblivious} reliable spanner is a distribution over spanners, and the bound on $|B^+|$ holds in expectation. We devise an oblivious $\nu$-reliable $(2+\frac{2}{k-1})$-spanner for any $k$-HST, whose lightness is $\approx \nu^{-2}$. We demonstrate a matching $\Omega(\nu^{-2})$ lower bound on the lightness (for any finite stretch). We also note that any stretch below 2 must incur linear lightness. For general metrics, doubling metrics, and metrics arising from minor-free graphs, we construct {\em light} tree covers, in which every tree is a $k$-HST of low weight. Combining these covers with our results for $k$-HSTs, we obtain oblivious reliable light spanners for these metric spaces, with nearly optimal parameters. In particular, for doubling metrics we get an oblivious $\nu$-reliable $(1+\varepsilon)$-spanner with lightness $\varepsilon^{-O({\rm ddim})}\cdot\tilde{O}(\nu^{-2}\cdot\log n)$, which is best possible (up to lower order terms).

cs.DS↗

arXiv2021

A Unified Framework for Hopsets and Spanners

Given an undirected graph $G=(V,E)$, an {\em $(\alpha,\beta)$-spanner} $H=(V,E')$ is a subgraph that approximately preserves distances; for every $u,v\in V$, $d_H(u,v)\le \alpha\cdot d_G(u,v)+\beta$. An $(\alpha,\beta)$-hopset is a graph $H=(V,E")$, so that adding its edges to $G$ guarantees every pair has an $\alpha$-approximate shortest path that has at most $\beta$ edges (hops), that is, $d_G(u,v)\le d_{G\cup H}^{(\beta)}(u,v)\le \alpha\cdot d_G(u,v)$. Given the usefulness of spanners and hopsets for fundamental algorithmic tasks, several different algorithms and techniques were developed for their construction, for various regimes of the stretch parameter $\alpha$. In this work we develop a single algorithm that can attain all state-of-the-art spanners and hopsets for general graphs, by choosing the appropriate input parameters. In fact, in some cases it also improves upon the previous best results. We also show a lower bound on our algorithm. In \cite{BP20}, given a parameter $k$, a $(O(k^{\epsilon}),O(k^{1-\epsilon}))$-hopset of size $\tilde{O}(n^{1+1/k})$ was shown for any $n$-vertex graph and parameter $0<\epsilon<1$, and they asked whether this result is best possible. We resolve this open problem, showing that any $(\alpha,\beta)$-hopset of size $O(n^{1+1/k})$ must have $\alpha\cdot \beta\ge\Omega(k)$.

cs.DS↗

arXiv2020

Improved Weighted Additive Spanners

Graph spanners and emulators are sparse structures that approximately preserve distances of the original graph. While there has been an extensive amount of work on additive spanners, so far little attention was given to weighted graphs. Only very recently [ABSKS20] extended the classical +2 (respectively, +4) spanners for unweighted graphs of size $O(n^{3/2})$ (resp., $O(n^{7/5})$) to the weighted setting, where the additive error is $+2W$ (resp., $+4W$). This means that for every pair $u,v$, the additive stretch is at most $+2W_{u,v}$, where $W_{u,v}$ is the maximal edge weight on the shortest $u-v$ path. In addition, [ABSKS20] showed an algorithm yielding a $+8W_{max}$ spanner of size $O(n^{4/3})$, here $W_{max}$ is the maximum edge weight in the entire graph. In this work we improve the latter result by devising a simple deterministic algorithm for a $+(6+\varepsilon)W$ spanner for weighted graphs with size $O(n^{4/3})$ (for any constant $\varepsilon>0$), thus nearly matching the classical +6 spanner of size $O(n^{4/3})$ for unweighted graphs. Furthermore, we show a $+(2+\varepsilon)W$ subsetwise spanner of size $O(n\cdot\sqrt{|S|})$, improving the $+4W_{max}$ result of [ABSKS20] (that had the same size). We also show a simple randomized algorithm for a $+4W$ emulator of size $\tilde{O}(n^{4/3})$. In addition, we show that our technique is applicable for very sparse additive spanners, that have linear size. For weighted graphs, we use a variant of our simple deterministic algorithm that yields a linear size $+\tilde{O}(\sqrt{n}\cdot W)$ spanner, and we also obtain a tradeoff between size and stretch. Finally, generalizing the technique of [DHZ00] for unweighted graphs, we devise an efficient randomized algorithm producing a $+2W$ spanner for weighted graphs of size $\tilde{O}(n^{3/2})$ in $\tilde{O}(n^2)$ time.

cs.DS↗

arXiv2020

Centralized, Parallel, and Distributed Multi-Source Shortest Paths via Hopsets and Rectangular Matrix Multiplication

Consider an undirected weighted graph $G = (V,E,w)$. We study the problem of computing $(1+\epsilon)$-approximate shortest paths for $S \times V$, for a subset $S \subseteq V$ of $|S| = n^r$ sources, for some $0 < r \le 1$. We devise a significantly improved algorithm for this problem in the entire range of parameter $r$, in both the classical centralized and the parallel (PRAM) models of computation, and in a wide range of $r$ in the distributed (Congested Clique) model. Specifically, our centralized algorithm for this problem requires time $\tilde{O}(|E| \cdot n^{o(1)} + n^{\omega(r)})$, where $n^{\omega(r)}$ is the time required to multiply an $n^r \times n$ matrix by an $n \times n$ one. Our PRAM algorithm has polylogarithmic time $(\log n)^{O(1/\rho)}$, and its work complexity is $\tilde{O}(|E| \cdot n^\rho + n^{\omega(r)})$, for any arbitrarily small constant $\rho >0$. In particular, for $r \le 0.313\ldots$, our centralized algorithm computes $S \times V$ $(1+\epsilon)$-approximate shortest paths in $n^{2 + o(1)}$ time. Our PRAM polylogarithmic-time algorithm has work complexity $O(|E| \cdot n^\rho + n^{2+o(1)})$, for any arbitrarily small constant $\rho >0$. Previously existing solutions either require centralized time/parallel work of $O(|E| \cdot |S|)$ or provide much weaker approximation guarantees. In the Congested Clique model, our algorithm solves the problem in polylogarithmic time for $|S| = n^r$ sources, for $r \le 0.655$, while previous state-of-the-art algorithms did so only for $r \le 1/2$. Moreover, it improves previous bounds for all $r > 1/2$. For unweighted graphs, the running time is improved further to $poly(\log\log n)$.

cs.DS↗

arXiv2020

Near-Additive Spanners and Near-Exact Hopsets, A Unified View

Given an {\em unweighted} undirected graph $G = (V,E)$, and a pair of parameters $\epsilon > 0$, $\beta = 1,2,\ldots$, a subgraph $G' =(V,H)$, $H \subseteq E$, of $G$ is a {\em $(1+\epsilon,\beta)$-spanner} (aka, a {\em near-additive spanner}) of $G$ if for every $u,v \in V$, $$d_{G'}(u,v) \le (1+\epsilon)d_G(u,v) + \beta~.$$ It was shown in \cite{EP01} that for any $n$-vertex $G$ as above, and any $\epsilon > 0$ and $\kappa = 1,2,\ldots$, there exists a $(1+\epsilon,\beta)$-spanner $G'$ with $O_{\epsilon,\kappa}(n^{1+1/\kappa})$ edges, with $$\beta = \beta_{EP} = \left({{\log \kappa} \over \epsilon}\right)^{\log \kappa - 2}~.$$ This bound remains state-of-the-art, and its dependence on $\epsilon$ (for the case of small $\kappa$) was shown to be tight in \cite{ABP18}. Given a {\em weighted} undirected graph $G = (V,E,\omega)$, and a pair of parameters $\epsilon > 0$, $\beta = 1,2,\ldots$, a graph $G'= (V,H,\omega')$ is a {\em $(1+\epsilon,\beta)$-hopset} (aka, a {\em near-exact hopset}) of $G$ if for every $u,v \in V$, $$d_G(u,v) \le d_{G\cup G'}^{(\beta)}(u,v) \le (1+\epsilon)d_G(u,v)~,$$ where $ d_{G\cup G'}^{(\beta)}(u,v)$ stands for a $\beta$-(hop)-bounded distance between $u$ and $v$ in the union graph $G \cup G'$. It was shown in \cite{EN16} that for any $n$-vertex $G$ and $\epsilon$ and $\kappa$ as above, there exists a $(1+\epsilon,\beta)$-hopset with $\tilde{O}(n^{1+1/\kappa})$ edges, with $\beta = \beta_{EP}$. Not only the two results of \cite{EP01} and \cite{EN16} are strikingly similar, but so are also their proof techniques. Moreover, Thorup-Zwick's later construction of near-additive spanners \cite{TZ06} was also shown in \cite{EN19,HP17} to provide hopsets with analogous (to that of \cite{TZ06}) properties. In this survey we explore this intriguing phenomenon, sketch the basic proof techniques used for these results, and highlight open questions.

cs.DS↗

arXiv2019

Almost Shortest Paths with Near-Additive Error in Weighted Graphs

Let $G=(V,E,w)$ be a weighted undirected graph with $n$ vertices and $m$ edges, and fix a set of $s$ sources $S\subseteq V$. We study the problem of computing {\em almost shortest paths} (ASP) for all pairs in $S \times V$ in both classical centralized and parallel (PRAM) models of computation. Consider the regime of multiplicative approximation of $1+\epsilon$, for an arbitrarily small constant $\epsilon > 0$ . In this regime existing centralized algorithms require $\Omega(\min\{|E|s,n^\omega\})$ time, where $\omega < 2.372$ is the matrix multiplication exponent. Existing PRAM algorithms with polylogarithmic depth (aka time) require work $\Omega(\min\{|E|s,n^\omega\})$. Our centralized algorithm has running time $O((m+ ns)n^\rho)$, and its PRAM counterpart has polylogarithmic depth and work $O((m + ns)n^\rho)$, for an arbitrarily small constant $\rho > 0$. For a pair $(s,v) \in S\times V$, it provides a path of length $\hat{d}(s,v)$ that satisfies $\hat{d}(s,v) \le (1+\epsilon)d_G(s,v) + \beta \cdot W(s,v)$, where $W(s,v)$ is the weight of the heaviest edge on some shortest $s-v$ path. Hence our additive term depends linearly on a {\em local} maximum edge weight, as opposed to the global maximum edge weight in previous works. Finally, our $\beta = (1/\rho)^{O(1/\rho)}$. We also extend a centralized algorithm of Dor et al. \cite{DHZ00}. For a parameter $\kappa = 1,2,\ldots$, this algorithm provides for {\em unweighted} graphs a purely additive approximation of $2(\kappa -1)$ for {\em all pairs shortest paths} (APASP) in time $\tilde{O}(n^{2+1/\kappa})$. Within the same running time, our algorithm for {\em weighted} graphs provides a purely additive error of $2(\kappa - 1) W(u,v)$, for every vertex pair $(u,v) \in {V \choose 2}$, with $W(u,v)$ defined as above. On the way to these results we devise a suit of novel constructions of spanners, emulators and hopsets.

cs.DS↗

arXiv2019

Lossless Prioritized Embeddings

Given metric spaces $(X,d)$ and $(Y,\rho)$ and an ordering $x_1,x_2,\ldots,x_n$ of $(X,d)$, an embedding $f: X \rightarrow Y$ is said to have a prioritized distortion $\alpha(\cdot)$, if for any pair $x_j,x'$ of distinct points in $X$, the distortion provided by $f$ for this pair is at most $\alpha(j)$. If $Y$ is a normed space, the embedding is said to have prioritized dimension $\beta(\cdot)$, if $f(x_j)$ may have nonzero entries only in its first $\beta(j)$ coordinates. The notion of prioritized embedding was introduced by \cite{EFN15}, where a general methodology for constructing such embeddings was developed. Though this methodology enables \cite{EFN15} to come up with many prioritized embeddings, it typically incurs some loss in the distortion. This loss is problematic for isometric embeddings. It is also troublesome for Matousek's embedding of general metrics into $\ell_\infty$, which for a parameter $k = 1,2,\ldots$, provides distortion $2k-1$ and dimension $O(k \log n \cdot n^{1/k})$. In this paper we devise two lossless prioritized embeddings. The first one is an isometric prioritized embedding of tree metrics into $\ell_\infty$ with dimension $O(\log j)$. The second one is a prioritized Matousek's embedding of general metrics into $\ell_\infty$, which provides prioritized distortion $2 \lceil k {{\log j} \over {\log n}} \rceil - 1$ and dimension $O(k \log n \cdot n^{1/k})$, again matching the worst-case guarantee $2k-1$ in the distortion of the classical Matousek's embedding. We also provide a dimension-prioritized variant of Matousek's embedding. Finally, we devise prioritized embeddings of general metrics into (single) ultra-metric and of general graphs into (single) spanning tree with asymptotically optimal distortion.

cs.DS↗

arXiv2019

Covering Metric Spaces by Few Trees

A {\em tree cover} of a metric space $(X,d)$ is a collection of trees, so that every pair $x,y\in X$ has a low distortion path in one of the trees. If it has the stronger property that every point $x\in X$ has a single tree with low distortion paths to all other points, we call this a {\em Ramsey} tree cover. Tree covers and Ramsey tree covers have been studied by \cite{BLMN03,GKR04,CGMZ05,GHR06,MN07}, and have found several important algorithmic applications, e.g. routing and distance oracles. The union of trees in a tree cover also serves as a special type of spanner, that can be decomposed into a few trees with low distortion paths contained in a single tree; Such spanners for Euclidean pointsets were presented by \cite{ADMSS95}. In this paper we devise efficient algorithms to construct tree covers and Ramsey tree covers for general, planar and doubling metrics. We pay particular attention to the desirable case of distortion close to 1, and study what can be achieved when the number of trees is small. In particular, our work shows a large separation between what can be achieved by tree covers vs. Ramsey tree covers.

cs.DS↗

arXiv2019

Distributed Construction of Light Networks

A $t$-{\em spanner} $H$ of a weighted graph $G=(V,E,w)$ is a subgraph that approximates all pairwise distances up to a factor of $t$. The {\em lightness} of $H$ is defined as the ratio between the weight of $H$ to that of the minimum spanning tree. An $(\alpha,\beta)$-{\em Shallow Light Tree} (SLT) is a tree of lightness $\beta$, that approximates all distances from a designated root vertex up to a factor of $\alpha$. A long line of works resulted in efficient algorithms that produce (nearly) optimal light spanners and SLTs. Some of the most notable algorithmic applications of light spanners and SLTs are in distributed settings. Surprisingly, so far there are no known efficient distributed algorithms for constructing these objects in general graphs. In this paper we devise efficient distributed algorithms in the CONGEST model for constructing light spanners and SLTs, with near optimal parameters. Specifically, for any $k\ge 1$ and $0<\epsilon<1$, we show a $(2k-1)\cdot(1+\epsilon)$-spanner with lightness $O(k\cdot n^{1/k})$ can be built in $\tilde{O}\left(n^{\frac12+\frac{1}{4k+2}}+D\right)$ rounds (where $n=|V|$ and $D$ is the hop-diameter of $G$). In addition, for any $\alpha>1$ we provide an $(\alpha,1+\frac{O(1)}{\alpha-1})$-SLT in $(\sqrt{n}+D)\cdot n^{o(1)}$ rounds. The running time of our algorithms cannot be substantially improved. We also consider spanners for the family of doubling graphs, and devise a $(\sqrt{n}+D)\cdot n^{o(1)}$ rounds algorithm in the CONGEST model that computes a $(1+\epsilon)$-spanner with lightness $(\log n)/\epsilon^{O(1)}$. As a stepping stone, which is interesting in its own right, we first develop a distributed algorithm for constructing nets (for arbitrary weighted graphs), generalizing previous algorithms that worked only for unweighted graphs.

cs.DS↗

arXiv2018

Light Spanners for High Dimensional Norms via Stochastic Decompositions

Spanners for low dimensional spaces (e.g. Euclidean space of constant dimension, or doubling metrics) are well understood. This lies in contrast to the situation in high dimensional spaces, where except for the work of Har-Peled, Indyk and Sidiropoulos (SODA 2013), who showed that any $n$-point Euclidean metric has an $O(t)$-spanner with $\tilde{O}(n^{1+1/t^2})$ edges, little is known. In this paper we study several aspects of spanners in high dimensional normed spaces. First, we build spanners for finite subsets of $\ell_p$ with $1<p\le 2$. Second, our construction yields a spanner which is both sparse and also {\em light}, i.e., its total weight is not much larger than that of the minimum spanning tree. In particular, we show that any $n$-point subset of $\ell_p$ for $1<p\le 2$ has an $O(t)$-spanner with $n^{1+\tilde{O}(1/t^p)}$ edges and lightness $n^{\tilde{O}(1/t^p)}$. In fact, our results are more general, and they apply to any metric space admitting a certain low diameter stochastic decomposition. It is known that arbitrary metric spaces have an $O(t)$-spanner with lightness $O(n^{1/t})$. We exhibit the following tradeoff: metrics with decomposability parameter $\nu=\nu(t)$ admit an $O(t)$-spanner with lightness $\tilde{O}(\nu^{1/t})$. For example, $n$-point Euclidean metrics have $\nu\le n^{1/t}$, metrics with doubling constant $\lambda$ have $\nu\le\lambda$, and graphs of genus $g$ have $\nu\le g$. While these families do admit a ($1+\epsilon$)-spanner, its lightness depend exponentially on the dimension (resp. $\log g$). Our construction alleviates this exponential dependency, at the cost of incurring larger stretch.

cs.DS↗

arXiv2018

Near Isometric Terminal Embeddings for Doubling Metrics

Given a metric space $(X,d)$, a set of terminals $K\subseteq X$, and a parameter $t\ge 1$, we consider metric structures (e.g., spanners, distance oracles, embedding into normed spaces) that preserve distances for all pairs in $K\times X$ up to a factor of $t$, and have small size (e.g. number of edges for spanners, dimension for embeddings). While such terminal (aka source-wise) metric structures are known to exist in several settings, no terminal spanner or embedding with distortion close to 1, i.e., $t=1+\epsilon$ for some small $0<\epsilon<1$, is currently known. Here we devise such terminal metric structures for {\em doubling} metrics, and show that essentially any metric structure with distortion $1+\epsilon$ and size $s(|X|)$ has its terminal counterpart, with distortion $1+O(\epsilon)$ and size $s(|K|)+1$. In particular, for any doubling metric on $n$ points, a set of $k=o(n)$ terminals, and constant $0<\epsilon<1$, there exists: (1) A spanner with stretch $1+\epsilon$ for pairs in $K\times X$, with $n+o(n)$ edges. (2) A labeling scheme with stretch $1+\epsilon$ for pairs in $K\times X$, with label size $\approx \log k$. (3) An embedding into $\ell_\infty^d$ with distortion $1+\epsilon$ for pairs in $K\times X$, where $d=O(\log k)$. Moreover, surprisingly, the last two results apply if only $K$ is a doubling metric, while $X$ can be arbitrary.

cs.DS↗

arXiv2017

Metric Embedding via Shortest Path Decompositions

We study the problem of embedding shortest-path metrics of weighted graphs into $\ell_p$ spaces. We introduce a new embedding technique based on low-depth decompositions of a graph via shortest paths. The notion of Shortest Path Decomposition depth is inductively defined: A (weighed) path graph has shortest path decomposition (SPD) depth $1$. General graph has an SPD of depth $k$ if it contains a shortest path whose deletion leads to a graph, each of whose components has SPD depth at most $k-1$. In this paper we give an $O(k^{\min\{\frac{1}{p},\frac{1}{2}\}})$-distortion embedding for graphs of SPD depth at most $k$. This result is asymptotically tight for any fixed $p>1$, while for $p=1$ it is tight up to second order terms. As a corollary of this result, we show that graphs having pathwidth $k$ embed into $\ell_p$ with distortion $O(k^{\min\{\frac{1}{p},\frac{1}{2}\}})$. For $p=1$, this improves over the best previous bound of Lee and Sidiropoulos that was exponential in $k$; moreover, for other values of $p$ it gives the first embeddings whose distortion is independent of the graph size $n$. Furthermore, we use the fact that planar graphs have SPD depth $O(\log n)$ to give a new proof that any planar graph embeds into $\ell_1$ with distortion $O(\sqrt{\log n})$. Our approach also gives new results for graphs with bounded treewidth, and for graphs excluding a fixed minor.

cs.DS↗

arXiv2017

Ramsey Spanning Trees and their Applications

The metric Ramsey problem asks for the largest subset $S$ of a metric space that can be embedded into an ultrametric (more generally into a Hilbert space) with a given distortion. Study of this problem was motivated as a non-linear version of Dvoretzky theorem. Mendel and Naor 2007 devised the so called Ramsey Partitions to address this problem, and showed the algorithmic applications of their techniques to approximate distance oracles and ranking problems. In this paper we study the natural extension of the metric Ramsey problem to graphs, and introduce the notion of Ramsey Spanning Trees. We ask for the largest subset $S\subseteq V$ of a given graph $G=(V,E)$, such that there exists a spanning tree of $G$ that has small stretch for $S$. Applied iteratively, this provides a small collection of spanning trees, such that each vertex has a tree providing low stretch paths to all other vertices. The union of these trees serves as a special type of spanner, a tree-padding spanner. We use this spanner to devise the first compact stateless routing scheme with $O(1)$ routing decision time, and labels which are much shorter than in all currently existing schemes. We first revisit the metric Ramsey problem, and provide a new deterministic construction. We prove that for every $k$, any $n$-point metric space has a subset $S$ of size at least $n^{1-1/k}$ which embeds into an ultrametric with distortion $8k$. This results improves the best previous result of Mendel and Naor that obtained distortion $128k$ and required randomization. In addition, it provides the state-of-the-art deterministic construction of a distance oracle. Building on this result, we prove that for every $k$, any $n$-vertex graph $G=(V,E)$ has a subset $S$ of size at least $n^{1-1/k}$, and a spanning tree of $G$, that has stretch $O(k \log \log n)$ between any point in $S$ and any point in $V$.

cs.DS↗

arXiv2017

Linear-Size Hopsets with Small Hopbound, and Distributed Routing with Low Memory

For a positive parameter $\beta$, the $\beta$-bounded distance between a pair of vertices $u,v$ in a weighted undirected graph $G = (V,E,\omega)$ is the length of the shortest $u-v$ path in $G$ with at most $\beta$ edges, aka {\em hops}. For $\beta$ as above and $\epsilon>0$, a {\em $(\beta,\epsilon)$-hopset} of $G = (V,E,\omega)$ is a graph $G' =(V,H,\omega_H)$ on the same vertex set, such that all distances in $G$ are $(1+\epsilon)$-approximated by $\beta$-bounded distances in $G\cup G'$. Hopsets are a fundamental graph-theoretic and graph-algorithmic construct, and they are widely used for distance-related problems in a variety of computational settings. Currently existing constructions of hopsets produce hopsets either with $\Omega(n \log n)$ edges, or with a hopbound $n^{\Omega(1)}$. In this paper we devise a construction of {\em linear-size} hopsets with hopbound $(\log n)^{\log^{(3)}n+O(1)}$. This improves the previous bound almost exponentially. We also devise efficient implementations of our construction in PRAM and distributed settings. The only existing PRAM algorithm \cite{EN16} for computing hopsets with a constant (i.e., independent of $n$) hopbound requires $n^{\Omega(1)}$ time. We devise a PRAM algorithm with polylogarithmic running time for computing hopsets with a constant hopbound, i.e., our running time is exponentially better than the previous one. Moreover, these hopsets are also significantly sparser than their counterparts from \cite{EN16}. We use our hopsets to devise a distributed routing scheme that exhibits near-optimal tradeoff between individual memory requirement $\tilde{O}(n^{1/k})$ of vertices throughout preprocessing and routing phases of the algorithm, and stretch $O(k)$, along with a near-optimal construction time $\approx D + n^{1/2 + 1/k}$, where $D$ is the hop-diameter of the input graph.

cs.DS↗