We address the All-Pairs Shortest Paths (APSP) problem for a number of unweighted, undirected geometric intersection graphs. We present a general reduction of the problem to static, offline intersection searching (specifically detection). As a consequence, we can solve APSP for intersection graphs of n arbitrary disks in O (n2 log n) time, axis-aligned line segments in O (n2 log log n) time, arbitrary line segments in O (n7/3 log1/3 n) time, d-dimensional axis-aligned boxes in O (n2 logd−1.5 n) time for d ≥ 2, and d-dimensional axis-aligned unit hypercubes in O (n2 log log n) time for d = 3 and O (n2 logd−3 n) time for d ≥ 4. In addition, we show how to solve the Single-Source Shortest Paths (SSSP) problem in unweighted intersection graphs of axis-aligned line segments in O (n log n) time, by a reduction to dynamic orthogonal point location.
CITATION STYLE
Chan, T. M., & Skrepetos, D. (2017). All-pairs shortest paths in geometric intersection graphs. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 10389 LNCS, pp. 253–264). Springer Verlag. https://doi.org/10.1007/978-3-319-62127-2_22
Mendeley helps you to discover research relevant for your work.