A dual simplex algorithm for the canonical representation of the uncapacitated facility location problem

Abstract

A number of algorithms have been designed for uncapacitated facility location problems. Computational experience with standard data sets seems to indicate the superiority of dual approaches, best known through efficient subgradient or dual adjustment procedures. However, linear programming algorithms provide distinct advantages, in particular direct amenability to cutting-plane techniques. A streamlined dual simplex algorithm is designed based on a covering formulation of the problem and present computational results.