A Simple Extension of Dirac's Theorem on Hamiltonicity

TL;DR

This paper extends Dirac's theorem to graphs with minimum degree at least half the vertices, identifying the only non-Hamiltonian cases, and provides a constructive polynomial-time algorithm for Hamiltonian cycle detection.

Contribution

It generalizes Dirac's theorem to a lower degree bound and introduces a polynomial-time algorithm for constructing Hamiltonian cycles in such graphs.

Findings

Identifies the only non-Hamiltonian graph families with minimum degree loor(n/2).

Provides a simple and an alternative constructive proof of the extended theorem.

Develops a polynomial-time algorithm for Hamiltonian cycle detection under the new conditions.

Abstract

The classical Dirac theorem asserts that every graph $G$ on $n$ vertices with minimum degree $\delta(G) \ge \lceil n/2 \rceil$ is Hamiltonian. The lower bound of $\lceil n/2 \rceil$ on the minimum degree of a graph is tight. In this paper, we extend the classical Dirac theorem to the case where $\delta(G) \ge \lfloor n/2 \rfloor $ by identifying the only non-Hamiltonian graph families in this case. We first present a short and simple proof. We then provide an alternative proof that is constructive and self-contained. Consequently, we provide a polynomial-time algorithm that constructs a Hamiltonian cycle, if exists, of a graph $G$ with $\delta(G) \ge \lfloor n/2 \rfloor$, or determines that the graph is non-Hamiltonian. Finally, we present a self-contained proof for our algorithm which provides insight into the structure of Hamiltonian cycles when $\delta(G) \ge \lfloor n/2 \rfloor$ and is promising for extending the results of this paper to the cases with smaller degree bounds.