We consider a fault tolerant version of the metric facility location problem in which every city, j, is required to be connected to rj facilities. We give the first non-trivial approximation algorithm for this problem, having an approximation guarantee of 3 Hk, where k is the maximum requirement and Hk is the k-th harmonic number. Our algorithm is along the lines of [2] for the generalized Steiner network problem. It runs in phases, and each phase, using a generalization of the primal-dual algorithm of [4] for the metric facility location problem, reduces the maximum residual requirement by 1.
CITATION STYLE
Jain, K., & Vazirani, V. V. (2000). An approximation algorithm for the fault tolerant metric facility location problem. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1913, pp. 177–182). Springer Verlag. https://doi.org/10.1007/3-540-44436-x_18
Mendeley helps you to discover research relevant for your work.