On the Small Quasi-kernel conjecture
Péter L. Erdős, Ervin Győri, Tamás Róbert Mezei, Nika Salia, Mykhaylo Tyomkyn
TL;DR
The paper surveys the theory of quasi-kernels in directed graphs, with a focus on the Small Quasi-kernel conjecture which asserts that any source-free digraph contains a quasi-kernel of size at most half the vertices. It analyzes the Chvátal-Lovász algorithm, its limitations, and constructions showing it need not produce small Qks, while also presenting reformulations and equivalences with the Kostochka-Luo-Shan conjecture. Key results include that KLS and the small Qk conjecture are equivalent, planarity-driven consequences via the four-color theorem, and a range of class-specific proofs; recent work extends to new bounds, kernel-like notions, and infinite-graph analogs. The survey also highlights open questions, structural conditions that guarantee small Qks, and the broader impact on understanding kernel-related questions in digraphs.
Abstract
An independent vertex subset $S$ of the directed graph $G$ is a kernel if the set of out-neighbors of $S$ is $V(G)\setminus S$. An independent vertex subset $Q$ of $G$ is a quasi-kernel if the union of the first and second out-neighbors contains $V(G)\setminus S$ as a subset. Deciding whether a directed graph has a kernel is an NP-hard problem. In stark contrast, each directed graph has quasi-kernel(s) and one can be found in linear time. In this article, we will survey the results on quasi-kernel and their connection with kernels. We will focus on the small quasi-kernel conjecture which states that if the graph has no vertex of zero in-degree, then there exists a quasi-kernel of size not larger than half of the order of the graph. The paper also contains new proofs and some new results as well.
