Multi-Path Algorithms for minimum-colour path problems with applications to approximating barrier resilience
Let G be a graph with zero or more colours assigned to its vertices, and let vs and vt be two vertices of G. The minimum-colour path problem is to determine the minimum over all vs–vt paths of the number of colours used, where a colour is considered used if it is assigned to any vertex in the path....
Saved in:
Published in | Theoretical computer science Vol. 553; pp. 74 - 90 |
---|---|
Main Authors | , |
Format | Journal Article |
Language | English |
Published |
Elsevier B.V
09.10.2014
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | Let G be a graph with zero or more colours assigned to its vertices, and let vs and vt be two vertices of G. The minimum-colour path problem is to determine the minimum over all vs–vt paths of the number of colours used, where a colour is considered used if it is assigned to any vertex in the path. Although this problem is NP-hard with strong hardness of approximation results, many problems can be formulated as instances of the minimum-colour path problem with additional constraints which may be exploited to allow polynomial-time solutions or close approximations. We introduce a family of approximation algorithms, referred to as the Multi-Path Algorithms, for minimum-colour path problems, and go on to show examples of constraints which would allow polynomial-time solutions or constant factor approximations.
In particular, we describe applications to variants of the barrier resilience problem: given a pair of points s and t and an arrangement A of n regions in the plane, the problem is to determine the minimum over all s–t paths of the number of regions intersected. We show how to reduce the barrier resilience problem to the minimum-colour path problem, and go on to show that the Multi-Path Algorithms guarantee a 1.5 approximation when regions are unit disks and s,t are separated by at least 23. |
---|---|
Bibliography: | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23 |
ISSN: | 0304-3975 1879-2294 |
DOI: | 10.1016/j.tcs.2014.04.009 |