English

Turing Kernelization for Finding Long Paths and Cycles in Restricted Graph Classes

Data Structures and Algorithms 2017-04-21 v2 Computational Complexity

Abstract

The NP-complete kk-Path problem asks whether a given undirected graph has a (simple) path of length at least kk. We prove that kk-Path has polynomial-size Turing kernels when restricted to planar graphs, graphs of bounded degree, claw-free graphs, or to K3,tK_{3,t}-minor-free graphs for some constant tt. This means that there is an algorithm that, given a kk-Path instance (G,k)(G,k) belonging to one of these graph classes, computes its answer in polynomial time when given access to an oracle that solves kk-Path instances of size polynomial in kk in a single step. The difficulty of kk-Path can therefore be confined to subinstances whose size is independent of the total input size, but is bounded by a polynomial in the parameter kk alone. These results contrast existing superpolynomial lower bounds for the sizes of traditional kernels for the kk-Path problem on these graph classes: there is no polynomial-time algorithm that reduces any instance (G,k)(G,k) to a single, equivalent instance (G,k)(G',k') of size polynomial in kk unless NPcoNP/polyNP \subseteq coNP/poly. The same positive and negative results apply to the kk-Cycle problem, which asks for the existence of a cycle of length at least kk. Our kernelization schemes are based on a new methodology called Decompose-Query-Reduce.

Keywords

Cite

@article{arxiv.1402.4718,
  title  = {Turing Kernelization for Finding Long Paths and Cycles in Restricted Graph Classes},
  author = {Bart M. P. Jansen},
  journal= {arXiv preprint arXiv:1402.4718},
  year   = {2017}
}

Comments

39 pages, 8 figures

R2 v1 2026-06-22T03:11:41.732Z