Minimum Spanning Tree of Imprecise Points Under $L_1$-metric

Abstract

Let $S$ be a set of imprecise points that is represented by axis-aligned pairwise disjoint squares in the plane. A precise instance of $S$ is a set of points, one from each region of $S$. In this paper, we study the optimal minimum spanning tree (textit{OptMST}) problem on $S$. The textit{OptMST} problem looks for the precise instance of $S$ such that the weight of the MST in this instance, maximize (Max-MST) or minimize (Min-MST) between all precise instances of~$S$ under $L_1$-metric. We present a $(frac{3}{7})$-approximation algorithm for Max-MST. This is an improvement on the best-known approximation factor of $1/3$. If $S$ satisfies $k$-separability property (the distance between any pair of squares are at least $k.a_{max}$ where $a_{max}$ is the maximum length of the squares), the factor parameterizes to $frac{2k+3}{2k+7}$. We propose a new lower bound for Min-MST problem on $S$ under $L_1$-metric where $S$ contains unit squares and provide an approximation algorithm with $(1+2sqrt{2})$ asymptotic factor.