Approximation algorithms for the single allocation problem in hub-and-spoke networks and related metric labeling problems

This paper deals with a single allocation problem in hub-and-spoke networks. We present a simple deterministic 3-approximation algorithm and randomized 2-approximation algorithm based on a linear relaxation problem and a randomized rounding procedure. We handle the case where the number of hubs is three, which is known to be NP-hard, and present a (5/4)-approximation algorithm. The single allocation problem includes a special class of the metric labeling problem, defined by introducing an assumption that both objects and labels are embedded in a common metric space. Under this assumption, we can apply our algorithms to the metric labeling problem without losing theoretical approximation ratios. As a byproduct, we also obtain a (4/3)-approximation algorithm for an ordinary metric labeling problem with three labels.

[1]  Tomomi Matsui,et al.  Linear Relaxation for Hub Network Design Problems , 2001 .

[2]  Horst W. Hamacher,et al.  Adapting polyhedral properties from facility to hub location problems , 2004, Discret. Appl. Math..

[3]  Rainer E. Burkard,et al.  Perspectives of Monge Properties in Optimization , 1996, Discret. Appl. Math..

[4]  Martine Labbé,et al.  A branch and cut algorithm for hub location problems with single assignment , 2005, Math. Program..

[5]  Tomomi Matsui,et al.  A study of the quadratic semi-assignment polytope , 2009, Discret. Optim..

[6]  Dimitris Bertsimas,et al.  On Dependent Randomized Rounding Algorithms , 1996, IPCO.

[7]  Sungsoo Park,et al.  European Journal of Operational Research a Linear Program for the Two-hub Location Problem , 2022 .

[8]  Jinhyeon Sohn,et al.  The single allocation problem in the interacting three-hub network , 2000, Networks.

[9]  Warren P. Adams,et al.  A Tight Linearization and an Algorithm for Zero-One Quadratic Programming Problems , 1986 .

[10]  Joseph Naor,et al.  A Linear Programming Formulation and Approximation Algorithms for the Metric Labeling Problem , 2005, SIAM J. Discret. Math..

[11]  Ravindra K. Ahuja,et al.  Network Flows , 2011 .

[12]  Morton E. O'Kelly,et al.  Hub‐and‐Spoke Networks in Air Transportation: An Analytical Review , 1999 .

[13]  Éva Tardos,et al.  Approximation algorithms for classification problems with pairwise relationships: metric labeling and Markov random fields , 2002, JACM.

[14]  Sungsoo Park,et al.  The single allocation problem in the interacting three-hub network , 2000 .

[15]  Dimitris Bertsimas,et al.  On dependent randomized rounding algorithms , 1999, Oper. Res. Lett..

[16]  Joseph Naor,et al.  The hardness of metric labeling , 2004, 45th Annual IEEE Symposium on Foundations of Computer Science.

[17]  Éva Tardos,et al.  Approximation algorithms for classification problems with pairwise relationships: metric labeling and Markov random fields , 1999, 40th Annual Symposium on Foundations of Computer Science (Cat. No.99CB37039).

[18]  M. O'Kelly,et al.  A quadratic integer program for the location of interacting hub facilities , 1987 .

[19]  Martin Skutella,et al.  Convex quadratic and semidefinite programming relaxations in scheduling , 2001, JACM.

[20]  Hiroo Saito,et al.  MATHEMATICAL ENGINEERING TECHNICAL REPORTS , 2004 .

[21]  James F. Campbell,et al.  Integer programming formulations of discrete hub location problems , 1994 .

[22]  T. Fujie,et al.  The Quadratic Semi-Assignment Polytope , 2004 .

[23]  Martine Labbé,et al.  Projecting the flow variables for hub location problems , 2004, Networks.