English

Planar 3-dimensional assignment problems with Monge-like cost arrays

Optimization and Control 2014-05-21 v1 Data Structures and Algorithms

Abstract

Given an n×n×pn\times n\times p cost array CC we consider the problem pp-P3AP which consists in finding pp pairwise disjoint permutations φ1,φ2,,φp\varphi_1,\varphi_2,\ldots,\varphi_p of {1,,n}\{1,\ldots,n\} such that k=1pi=1nciφk(i)k\sum_{k=1}^{p}\sum_{i=1}^nc_{i\varphi_k(i)k} is minimized. For the case p=np=n the planar 3-dimensional assignment problem P3AP results. Our main result concerns the pp-P3AP on cost arrays CC that are layered Monge arrays. In a layered Monge array all n×nn\times n matrices that result from fixing the third index kk are Monge matrices. We prove that the pp-P3AP and the P3AP remain NP-hard for layered Monge arrays. Furthermore, we show that in the layered Monge case there always exists an optimal solution of the pp-3PAP which can be represented as matrix with bandwidth 4p3\le 4p-3. This structural result allows us to provide a dynamic programming algorithm that solves the pp-P3AP in polynomial time on layered Monge arrays when pp is fixed.

Keywords

Cite

@article{arxiv.1405.5210,
  title  = {Planar 3-dimensional assignment problems with Monge-like cost arrays},
  author = {Ante Ćustić and Bettina Klinz and Gerhard J. Woeginger},
  journal= {arXiv preprint arXiv:1405.5210},
  year   = {2014}
}

Comments

16 pages, appendix will follow in v2

R2 v1 2026-06-22T04:19:19.475Z