On dominating pair degree conditions for hamiltonicity in balanced bipartite digraphs

https://doi.org/10.1016/j.disc.2020.112240Get rights and content

Abstract

We prove several new sufficient conditions for hamiltonicity and bipancyclicity in balanced bipartite digraphs, in terms of sums of degrees over dominating or dominated pairs of vertices.

Introduction

This article is concerned with sufficient conditions for hamiltonicity and bipancyclicity in balanced bipartite digraphs. More specifically, we study several Meyniel-type criteria, that is, theorems asserting existence of hamiltonian cycles under certain conditions on the sums of degrees of non-adjacent vertices. There are numerous such criteria, and open problems, in general digraphs (see, e.g., [4], [5] and the references therein). Over the last few years, various analogues of these theorems and conjectures have been established for bipartite digraphs [1], [2], [3], [7], [8], [10], [11]. These results, generally speaking, do not follow from their non-bipartite analogues and require different arguments and techniques.

We begin with a short review of the relevant results to provide context for our present work. Throughout this paper, D denotes a strongly connected balanced bipartite digraph of order 2a (see Section 2 for details on notation and terminology).

The main feature of Meyniel-type criteria is that a degree condition be only imposed on pairs of non-adjacent vertices. The first such criterion in the bipartite setting was proved in [3].

Theorem 1.1

Let D be as above, with a2, and suppose that d(u)+d(v)3afor every pair of distinct vertices {u,v} such that uvA(D) and vuA(D). Then, D is hamiltonian.

The lower bound of 3a is sharp (see examples in [3]). The condition from Theorem 1.1 may be further strengthened, in the spirit of [5], by requiring that it be satisfied only by dominating and dominated pairs of vertices. This was done in [1].

Theorem 1.2

Let D be as above, with a3, and suppose that d(u)+d(v)3awhenever {u,v} is a dominating or dominated pair. Then, D is hamiltonian.

At this point, there are two natural questions: First, are the above assumptions enough to imply existence of cycles of all even lengths in D, perhaps modulo some exceptional digraphs (Bondy’s metaconjecture)? And secondly, could we expect the same conclusion if the degree sum condition was only satisfied by the dominating pairs of vertices? The answer to the first question is positive. More precisely, we have the following result.

Theorem 1.3

Let D be as above, with a3, and suppose that d(u)+d(v)3awhenever {u,v} is a dominating or dominated pair. Then, D is either bipancyclic or a directed cycle of length 2a.

The second question seems much harder. However, most recently, Wang and Wu [11] proposed an interesting variant of the degree sum condition that allows them to obtain hamiltonicity by only imposing the condition on dominating pairs of vertices.

Theorem 1.4

Let D be a strongly connected balanced bipartite digraph of order 2a, where a3, and let k be an integer satisfying max{1,a4}<ka2. Suppose that for every dominating pair {u,v} of vertices in D, d(u)2akandd(v)a+k,ord(u)a+kandd(v)2ak.Then, D is hamiltonian.

The authors of [11] posed also several interesting problems related to the above theorems. Among them:

  • (a)

    Are the assumptions of Theorem 1.4 enough to imply bipancyclicity of D?

  • (b)

    Is there an integer k0 such that D is hamiltonian if the inequality d(u)+d(v)3a+k is only imposed on the dominating pairs {u,v}?

The main goal of the present article is to prove the following positive answers to these two questions.

Theorem 1.5

If D satisfies the hypotheses of Theorem 1.4, then D is either bipancyclic or a directed cycle of length 2a.

Theorem 1.6

Let D be a strongly connected balanced bipartite digraph of order 2a, where a2. Suppose that for every dominating pair {u,v} of vertices in D, d(u)+d(v)3a+1.Then, D is hamiltonian.

Theorem 1.5, Theorem 1.6 are proved in Sections 4 Proof of, 3 Proof of, respectively. In the last section, we discuss some corollaries and open problems.

Section snippets

Notation and terminology

We consider digraphs in the sense of [4]: A digraph D is a pair (V(D),A(D)), where V(D) is a finite set (of vertices) and A(D) is a set of ordered pairs of distinct elements of V(D), called arcs (i.e., D has no loops or multiple arcs).

The number of vertices |V(D)| is the order of D (also denoted by |D|). For vertices u and v from V(D), we write uvA(D) to say that A(D) contains the ordered pair (u,v). If uvA(D), then u is called an in-neighbour of v, and v is an out-neighbour of u. A pair of

Proof of Theorem 1.6

Throughout this section we assume that D is a strongly connected balanced bipartite digraph with partite sets of cardinalities a2, which satisfies condition (D1). The proof of Theorem 1.6 is based on the following four simple lemmas.

Lemma 3.1

Suppose that D is non-hamiltonian. Then, for every vertex uV(D) there exists a vertex vV(D){u} such that {u,v} is a dominating pair.

Proof

For a proof by contradiction, suppose that D contains a vertex u0 which has no common out-neighbour with any other vertex in D. We

Proof of Theorem 1.5

The proof of Theorem 1.5 is based on Theorem 1.3, Theorem 1.4, and the following result of Thomassen.

Theorem 4.1

[9, Thm. 3.5]

Let G be a strongly connected digraph of order n, n3, such that d(u)+d(v)2n whenever u and v are non-adjacent. Then, G is either pancyclic, or a tournament, or n is even and G is isomorphic to Kn2,n2.

Let a,k be integers such that a3 and max{1,a4}<ka2. Note that then a+k2ak. Throughout this section, we assume that D is a strongly connected balanced bipartite digraph of order 2a, which

Final remarks

First of all, for the sake of completeness, let us note that analogues of Theorem 1.5, Theorem 1.6 for dominated pairs hold true as well. More precisely, we have the following results.

Theorem 5.1

Let D be a strongly connected balanced bipartite digraph of order 2a, where a3, and let k be an integer satisfying max{1,a4}<ka2. Suppose that for every dominated pair {u,v} of vertices in D, d(u)2akandd(v)a+k,ord(u)a+kandd(v)2ak.Then, D is either bipancyclic or a directed cycle of length 2a.

Theorem 5.2

Let D be a

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References (11)

There are more references available in the full text version of this article.

The research was partially supported by Natural Sciences and Engineering Research Council of Canada .

View full text