3 D iscrete N etw ork L ocation M o d e ls John Current 1 , Mark Daskin 2 , and David Schilling 1 1 D epartm ent of M anagem ent Sciences, F isher C ollege of B usiness, T he O hio State U niversity,2100 N eilA venue,C olum bus,O hio 43210. 2 D epartm ent of Industrial E ngineering and M anagem ent Sciences, N orthw estern U niversity,2145 Sheridan R oad,E vanston,IL 60208. 3 .1 In t r o d u c t io n Undoubtedly, humans have been analyzing the e®ectiveness of locational de- cisions since they inhabited their ¯rst cave. We use the term \facility" here in its broadest sense. That is, it is meant to include entities such as air and mar- itime ports, factories, warehouses, retail outlets, schools, hospitals, day-care centers, bus stops, subway stations, electronic switching centers, computer concentrators and terminals, rain gages, emergency warning sirens, and satel- lites, to name but a few that have been analyzed in the research literature. The ubiquity of locational decision-making has led to a strong interest in location analysis and modeling within the operations research and man- agement science community. The long and voluminous history of location research results from several factors. First, location decisions are frequently made at all levels of human organization from individuals and households to ¯rms, government agencies and even international agencies. Second, such decisions are often strategic in nature. That is, they involve large sums of cap- ital resources and their economic e®ects are long term. In the private sector they have a major in°uence on the ability of a ¯rm to compete in the market place. In the public sector they in°uence the e±ciency by which jurisdictions provide public services and the ability of these jurisdictions to attract house- holds and other economic activity. Third, they frequently impose economic externalities. Such externalities include pollution, congestion, and economic development, among others. Fourth, location models are often extremely di±cult to solve, at least optimally. Even the most basic models are computationally intractable for large problem instances. In fact, the computational complexity of location models is a major reason that the widespread interest in formulating and implementing such models did not occur until the advent of high-speed digital computers. Finally, location models are application speci¯c. That is, their structural form (the objectives, constraints and variables) is determined by the particular location problem under study. Consequently, there does not exist a general location model that is appropriate for all potential or existing applications.