Solving a Class of Two-Dimensional Uncapacitated Location-Allocation Problems by Dynamic Programming

In this paper we define and analyze a class of two-dimensional location-allocation problems that can be solved with a one-dimensional dynamic programming algorithm. We define a criterion that must be satisfied in order that a problem can be classified as having a one-dimensional intrinsic property. An algorithm is developed to test any given problem to see if it possesses this property. We then show that any problem possessing the intrinsic property can be solved by means of an efficient dynamic programming algorithm developed earlier by one of the authors.