Approximating Generalized Network Design under (Dis)economies of Scale with Applications to Energy Efficiency

Abstract

In a generalized network design (GND) problem, a set of resources are assigned (non-exclusively) to multiple requests . Each request contributes its weight to the resources it uses and the total load on a resource is then translated to the cost it incurs via a resource-specific cost function. Motivated by energy efficiency applications, recently, there is a growing interest in GND using cost functions that exhibit (dis)economies of scale ((D)oS) , namely, cost functions that appear subadditive for small loads and superadditive for larger loads. The current article advances the existing literature on approximation algorithms for GND problems with (D)oS cost functions in various aspects: (1) while the existing results are restricted to routing requests in undirected graphs, identifying the resources with the graph’s edges, the current article presents a generic approximation framework that yields approximation results for a much wider family of requests (including various types of Steiner tree and Steiner forest requests) in both directed and undirected graphs, where the resources can be identified with either the edges or the vertices; (2) while the existing results assume that a request contributes the same weight to each resource it uses, our approximation framework allows for unrelated weights, thus providing the first non-trivial approximation for the problem of scheduling unrelated parallel machines with (D)oS cost functions; (3) while most of the existing approximation algorithms are based on convex programming, our approximation framework is fully combinatorial and runs in strongly polynomial time; (4) the family of (D)oS cost functions considered in the current article is more general than the one considered in the existing literature, providing a more accurate abstraction for practical energy conservation scenarios; and (5) we obtain the first approximation ratio for GND with (D)oS cost functions that depends only on the parameters of the resources’ technology and does not grow with the number of resources, the number of requests, or their weights. The design of our approximation framework relies heavily on Roughgarden’s smoothness toolbox [43], thus demonstrating the possible usefulness of this toolbox in the area of approximation algorithms.