Minimal spectral radius of graphs with given matching number
Jiaqi Liu, Zhenzhen Lou, Vilmar Trevisan
TL;DR
This work resolves the Brualdi-Hoffman problem for graphs with fixed matching number by proving that spectrally minimal graphs in $\mathcal{G}_{n,\beta}$ are necessarily trees and by developing a unifying quasi-adjacency framework to classify such trees. It shows that a size-$\beta$ dominating set with controlled quasi-adjacency properties governs the minimizers and then derives explicit minimizers for $\beta=2,3,4$ using constructive parametrizations and existing results. The paper leverages spectral comparison techniques, spanning-tree preservation of the matching number, and Perron-vector-based comparisons to exclude non-minimizers. The resulting explicit minimizers provide concrete, case-based formulas, advancing the understanding of extremal spectral graph theory under a fixed matching constraint and offering precise templates for constructing extremal trees.
Abstract
The Brualdi-Solheid problem asks which graph achieves the extremal (maximum or minimum) spectral radius for a given class of graphs. This paper addresses the Brualdi-Solheid problem for \( \mathcal{G}_{n,β} \), the family of graphs with order \( n \) and matching number \( β\), aiming to identify its spectrally minimal graphs i.e., those that minimize the spectral radius \(ρ(G)\). We introduce the novel concept of ``quasi-adjacency'' relation, developing a unified structural classification framework for trees in \(\mathcal{G}_{n,β}\), which clarifies structural properties and provides a constructive method to generate trees with fixed \(β\). By showing that all spectrally minimal graphs in \( \mathcal{G}_{n,β} \) are trees, we further narrow the search for extremal graphs. Additionally, we apply this framework to the representative cases \(β=2,3,4\), obtaining the minimizers by explicit structural formulas involving parameters related to \(n\).
