English

Spanning trees in randomly perturbed graphs

Combinatorics 2018-03-14 v1

Abstract

A classical result of Koml\'os, S\'ark\"ozy and Szemer\'edi states that every nn-vertex graph with minimum degree at least (1/2+o(1))n(1/2+ o(1))n contains every nn-vertex tree with maximum degree O(n/logn)O(n/\log{n}) as a subgraph, and the bounds on the degree conditions are sharp. On the other hand, Krivelevich, Kwan and Sudakov recently proved that for every nn-vertex graph GαG_\alpha with minimum degree at least αn\alpha n for any fixed α>0\alpha >0 and every nn-vertex tree TT with bounded maximum degree, one can still find a copy of TT in GαG_\alpha with high probability after adding O(n)O(n) randomly-chosen edges to GαG_\alpha. We extend their results to trees with unbounded maximum degree. More precisely, for a given no(1)Δcn/lognn^{o(1)}\leq \Delta\leq cn/\log n and α>0\alpha>0, we determine the precise number (up to a constant factor) of random edges that we need to add to an arbitrary nn-vertex graph GαG_\alpha with minimum degree αn\alpha n in order to guarantee a copy of any fixed nn-vertex tree TT with maximum degree at most~Δ\Delta with high probability.

Keywords

Cite

@article{arxiv.1803.04958,
  title  = {Spanning trees in randomly perturbed graphs},
  author = {Felix Joos and Jaehoon Kim},
  journal= {arXiv preprint arXiv:1803.04958},
  year   = {2018}
}

Comments

41 pages

R2 v1 2026-06-23T00:52:01.542Z