Approximation algorithms for bounded facility location

Abstract

The bounded k-median problem is to select in an undirected graph G = (V,E) a set S of k vertices such that the maximum distance from a vertex υ ∈ V to S is at most a given bound d and the average distance from vertices V to S is minimized. We present randomized algorithms for several versions of this problem. We also study the bounded version of the uncapacitated facility location problem. For this latter problem we present extensions of known deterministic algorithms for the unbounded version, and we prove some inapproximability results.