Coverage Location Models: Alternatives, Approximation, and Uncertainty. (January 2016)

• Record Type: Journal Article
• Authors: Wei, Ran
• Other Names: Murray Alan T. guest-editor.; Church Richard L. guest-editor.
• Is Part Of: International regional science review. Volume 39:Number 1(2016)
• Page Start: 48
• Page End: 76
• Publication Date: 2016-01
• Subjects: maximal coverage -- continuous space -- discretization; Regional economics -- Periodicals; Regional planning -- Periodicals; Regionalism -- Periodicals; 338.9
• DOI: 10.1177/0160017615571588
• Languages: English
• ISSNs: 1552-6925

Abstract

Achieving maximal coverage of service facilities has been of great interest to urban and regional planners. Examples include placing cellular towers, siting emergency response stations, and locating weather radars, among others. In some planning contexts, facilities could be sited almost anywhere in a region due to their small geographic footprints and demand is continuously distributed. This location problem has been represented as the continuous space maximal coverage problem (CSMCP). The CSMCP is widely acknowledged to be challenging to solve exactly. A broadly used solution approach for the CSMCP is to transform the problem into discrete maximal coverage models through continuous space discretization. A variety of discrete simplifications of CSMCP have been developed, attempting to address spatial representation issues that arise in the application of discrete models used as a continuous space approximation. However, the performance of applying these discrete coverage models to approximately solving the CSMCP has not been explicitly evaluated. It remains elusive as to which approach provides the best approximation for the CSMCP. This article therefore presents a comparative performance analysis of various discrete approximations for the CSMCP. Empirical results provide insights on how to achieve a balance between model representation detail and reasonable computation. Potential research directions are also suggested.