A graph G(V, E) (|V| 2k) satisfies property A k if, given k pairs of distinct nodes (s 1 , t 1 ), ..., (s k , t k ) of V(G), there are k mutually node-disjoint paths, one connecting s i and t i for each i, 1 i k. A necessary condition for any graph to satisfy A k is that it is (2k&1)-connected. Hype
Efficient Algorithms for Finding the Maximum Number of Disjoint Paths in Grids
✍ Scribed by Wun-Tat Chan; Francis Y.L. Chin
- Publisher
- Elsevier Science
- Year
- 2000
- Tongue
- English
- Weight
- 366 KB
- Volume
- 34
- Category
- Article
- ISSN
- 0196-6774
No coin nor oath required. For personal study only.
✦ Synopsis
In a rectangular grid, given two sets of nodes, S S sources and T T sinks , of size 2 Ž . each, the disjoint paths DP problem is to connect as many nodes in S S to the Ž nodes in T T using a set of ''disjoint'' paths. Both edge-disjoint and ¨ertex-disjoint . cases are considered in this paper. Note that in this DP problem, a node in S S can be connected to any node in T T. Although in general the sizes of S S and T T do not have to be the same, algorithms presented in this paper can also find the maximum number of disjoint paths pairing nodes in S S and T T. We use the network flow approach to solve this DP problem. By exploiting all the properties of the network, such as planarity and regularity of a grid, integral flow, and unit capacity Ž sourcersinkrflow, we can optimally compress the size of the working grid to be . Ž 2 . Ž 1.5 . Ž 2.5 . defined from O N to O N and solve the problem in O N time for both the edge-disjoint and vertex-disjoint cases, an improvement over the straightfor-Ž 3 . ward approach which takes O N time.
📜 SIMILAR VOLUMES