Approximating Minimum Cost Source Location Problems with Local Vertex-Connectivity Demands

The source location problem is a problem of computing a minimum cost source set in an undirected graph so that the connectivity between the source set and a vertex is at least the demand of the vertex. In this paper, the connectivity between a source set S and a vertex v is defined as the maximum number of paths between v and S no two of which have common vertex except v. We propose an O(d∗ log d∗)-approximation algorithm for the problem with maximum demand d∗. We also define a variant of the source location problem and propose an approximation algorithm for it.