Approximation Schemes for the Generalized Traveling Salesman Problem
- 34 Downloads
The Generalized Traveling Salesman Problem (GTSP) is defined by a weighted graph G = (V,E,w) and a partition of its vertex set into k disjoint clusters V = V1 ∪... ∪ Vk. It is required to find a minimum-weight cycle that contains exactly one vertex of each cluster. We consider a geometric setting of the problem (we call it the EGTSP-k-GC), in which the vertices of the graph are points in the plane, the weight function corresponds to the Euclidean distances between the points, and the partition into clusters is specified implicitly by means of a regular integer grid with step 1. In this setting, a cluster is a subset of vertices lying in the same cell of the grid; the arising ambiguity is resolved arbitrarily. Even in this special setting, the GTSP remains intractable, generalizing in a natural way the classical planar Euclidean TSP. Recently, a \((1.5 + 8\sqrt 2 + \varepsilon )\) -approximation algorithm with complexity depending polynomially both on the number of vertices n and on the number of clusters k has been constructed for this problem. We propose three approximation schemes for this problem. For each fixed k, all the schemes are polynomial and the complexity of the first two is linear in the number of nodes. Furthermore, the first two schemes remain polynomial for k = O(log n), whereas the third scheme is polynomial for k = n − O(log n).
Keywordsgeneralized traveling salesman problem NP-hard problem polynomial-time approximation scheme.
Unable to display preview. Download preview PDF.
- 4.B. Bhattacharya, A. ´Custi´c, A. Rafiey, A. Rafiey, and V. Sokol, “Approximation algorithms for generalized MST and TSP in grid clusters,” in Combinatorial Optimization and Applications: Proceedings of the 9th International Conference, Houston, TX, USA, 2015 (Springer, Cham, 2015), pp. 110–125.CrossRefGoogle Scholar
- 7.N. Christofides, Worst-Case Analysis of a New Heuristic for the Traveling Salesman Problem, Technical Report No. AD-A025 602 (Carnegie Mellon Univ. Pittsburgh, 1976).Google Scholar
- 9.A. Dumitrescu and J. S. B. Mitchell, “Approximation algorithms for TSP with neighborhoods in the plane,” in Proceedings of the 12th Annual ACM–SIAM Symposium on Discrete Algorithms, Washington, DC, USA, 2001 (SIAM, Philadelphia, 2001), pp. 38–46.Google Scholar
- 10.A. Dumitrescu and C. D. Tóth, “The traveling salesman problem for lines, balls, and planes,” ACM Trans. Algorithms 12 (3), article 43 (2016).Google Scholar
- 17.C. S. Mata and J. S. B. Mitchell, “Approximation algorithms for geometric tour and network design problems,” in Proceedings of the 11th Annual Symposium on Computational Geometry, Vancouver, Canada, 1995 (ACM, New York, 1995), pp. 360–369.Google Scholar
- 18.J. S. B. Mitchell, “A PTAS for TSP with neighborhoods among fat regions in the plane,” in Proceedings of the 18th Annual ACM–SIAM Symposium on Discrete Algorithms, New Orleans, LA, USA, 2007 (SIAM, Philadelphia, 2007), pp. 11–18.Google Scholar