Improved Guarantees for Vertex Sparsification in Planar Graphs

Graphs Sparsification aims at compressing large graphs into smaller ones while (approximately) preserving important characteristics of the input graph. In this work we study Vertex Sparsifiers, i.e., sparsifiers whose goal is to reduce the number of vertices. More concretely, given a weighted graph $G=(V,E)$, and a terminal set $K$ with $|K|=k$, a quality $q$ vertex cut sparsifier of $G$ is a graph $H$ such that $K\subseteq V(H)$ and for any bipartition $U,K \setminus U$ of the terminal set $K$, the values of minimum cut separating $U$ from $K \setminus U$ in $G$ and $H$ are within a factor $q$ from each other. Similarly, we define vertex flow and distance sparsifiers that (approximately) preserve multicommodity flows and distances among terminal pairs, respectively. We study such vertex sparsifiers for planar graphs. We show that if all the $k$ terminals in $K$ lie on the same face of the input planar graph $G$, then there exist quality $1$ vertex cut, flow and distance sparsifiers of size $O(k^2)$ that are also planar. This improves upon the previous best known bound $O(k^22^{2k})$ for cut and flow sparsifiers for such class of graphs by an exponential factor. Our upper bound for cut sparsifiers also matches the known lower bound $\Omega(k^2)$ on the number of edges of such sparsifiers for this class of $k$-terminal planar graphs. We also provide a new lower bound of $\Omega(k^{1+1/(t-1)})$ on the size of any data structure that approximately preserves the pairwise terminal distance in sparse \emph{general} graphs within a multiplicative factor of $t$ or an additive error $2t-3$, for any $t\geq 2$.