Factors and connected factors in tough graphs with high isolated toughness
Morteza Hasanvand

TL;DR
This paper establishes conditions under which graphs with high isolated toughness contain specific factors and connected factors, extending understanding of graph structure related to toughness and factorization.
Contribution
It introduces new sufficient conditions for the existence of particular factors and connected factors in graphs with high isolated toughness.
Findings
Graphs satisfying certain isolated toughness conditions have specific factors.
Conditions for the existence of factors with prescribed degrees are established.
Connected factors with degrees in a specified set exist under given toughness constraints.
Abstract
Let be a graph and let be a positive integer-valued function on . Assume that for all , where denotes the set of isolated vertices of . In this paper, we show that if for all , and is even, then has a factor such that for each vertex , , where denotes the number of components of . Moreover, we show that if for all , and , then has a connected factor such that for each vertex , .
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAdvanced Graph Theory Research · Graph theory and applications · Graph Labeling and Dimension Problems
Factors and connected factors in tough graphs with high isolated toughness
Morteza Hasanvand
Department of Mathematical Sciences, Sharif University of Technology, Tehran, Iran
Abstract
Let be a graph and let be a positive integer-valued function on . Assume that for all ,
[TABLE]
where denotes the set of isolated vertices of . In this paper, we show that if for all ,
[TABLE]
and is even, then has a factor such that for each vertex , , where denotes the number of components of . Moreover, we show that if for all ,
[TABLE]
and , then has a connected factor such that for each vertex , .
*Keywords:
Toughness; isolated toughness; regular factor; connected factor; -factor. *
1 Introduction
In this article, all graphs have no loop, but multiple edges are allowed and a simple graph is a graph without multiple edges. Let be a graph. The vertex set and the edge set of are denoted by and , respectively. We also denote by , , and the number of isolated vertices of , the number of components of with odd number of vertices, and the number of components of , respectively. For a vertex set of , we denote by the induced subgraph of with the vertex set containing precisely those edges of whose ends lie in . The vertex set is called an independent set, if there is no edge of connecting vertices in . The maximum size of all independent sets of is denoted by . Let be a positive real number, a graph is said to be -tough, if for all . Furthermore, is said to be -iso-tough, if for all . This definition is a little different from [13, 14] for the sake of simplicity. Note that when is -iso-tough, for each vertex , the number of its neighbours must be at least and hence the conditions and must automatically hold. More generally, when is a real function on , we say that is -iso-tough, if for all , , where denotes the set of all isolated vertices of . We denote by the set of all neighbours of vertices of in . For a set of integers, an -factor is a spanning subgraph with vertex degrees in . Let and be two integer-valued functions on . A -factor of is a spanning subgraph such that for each vertex , . When , we call it a -factor as well. An -factor of refers to a spanning subgraph such that for each vertex , . A near -factor refers to a spanning subgraph such that for each vertex , , except for at most one vertex with . Note that when the sum of all taken over all vertices is even, is a near -factor if and only if is an -factor. Note that several theorems in graph theory for the existence of -factors can be developed to a near -factor version. This type of factors is useful when a factor is required for extending to connected factors with bounded degrees as the proof of Theorem 4.8. For example, see [3, 9]. For convenience, we write for and write for . Let and be two disjoint vertex sets. We denote by the number of components of satisfying , where denotes the number of edges of with one end in and the other one in . Throughout this article, all variables are positive integers.
In 1947 Tutte introduced the following criterion for the existence of a perfect matching.
Theorem 1.1
.([16])* A graph has a -factor if and only if for all , .*
In 1978 Vergenas formulated a criterion for the existence of -factors and showed that the criterion becomes simpler for the following special case.
Theorem 1.2
.([11])* Let be a graph and let be an integer-valued function on with . Then has a -factor if and only if for all , *
In 1985 Enomoto, Jackson, Katerinis, and Saito proved the following theorem on tough graphs, which was originally conjectured by Chvátal (1973) [5]. In 1990 Katerinis [10] generalized their result by replacing a weaker sufficient toughness condition for the existence of -factors, provided that .
Theorem 1.3
.([8])* Every -tough graph of order at least with even has a -factor.*
In 2007 Ma and Yu strengthened Katerinis’ result by replacing isolated toughness condition as the following theorem. In this paper, we provide a supplement for their result by improving Theorem 1.3 for -iso-tough graphs by showing that the needed toughness can be pushed down to the fixed number . In Section 5, we also establish another refined version in -iso-tough graphs.
Theorem 1.4
.([13])* Every -iso-tough graph has an -factor, when .*
In 1973 Chvátal [5] conjectured that there exists a positive real number such that every -tough graph of order at least three admits a Hamiltonian cycle. In 2000 Ellingham and Zha [7] confirmed a weaker version of this conjecture by proving that every -tough graph of order at least three admits a connected -factor. Motivated by this result, one way ask whether higher toughness can guarantee the existence of connected -factors. The following theorem shows that the answer is positive. In this paper, we show that the needed toughness of this theorem can be pushed down to the fixed number but in -iso-tough graphs.
Theorem 1.5
.([6, 7], see [9])* Every -tough graph of order at least has a connected -factor, where .*
In 1990 Katerinis formulated the following sufficient toughness condition for the existence of -factors. In Section 4, we introduce some sufficient toughness conditions for the existence of -factors and connected -factors in graphs with high enough isolated toughness as mentioned in the abstract.
Theorem 1.6
.([10])* Let be a graph and let be a positive integer-valued function on satisfying , where and are two positive integers. If is -tough and is even, then has an -factor.*
2 Tools and preliminary results
In this section, we shall provide some necessary tools for applying in the next sections. Before doing so, let us recall a theorem due to Tutte (1952) as the following version.
Theorem 2.1
.([17])* Let be a general graph and let be an integer-valued function on . Then has a near -factor if and only if for all disjoint subsets and of ,*
[TABLE]
The following corollary is an application of Theorem 2.1, which is inspired by Lemma 4 in [10].
Corollary 2.2
.(see [10])* Let be a general graph and let be an integer-valued function on . Then has a near -factor, if*
[TABLE]
for all disjoint subsets and of satisfying and for each .
**Proof. **
Let us define . We are going to show that the inequality holds for any two disjoint subsets and of and so the proof follows from Theorem 2.1. By induction on . Let be the right-hand side of the inequality in the corollary. Assume that has a vertex with or . Define . If , then
[TABLE]
where denotes the number of edges of incident to with the other end in . Also, if , then
[TABLE]
where . Hence the lemma holds.
- *
The following theorem can generalize Lemma 1 in [10] and plays an essential role in this paper.
Theorem 2.3
.* Let be a graph. If is a nonnegative real function on , then there is a maximal independent subset of such that*
[TABLE]
**Proof. **
Define . For every nonnegative integer with , take to be a vertex of with the maximum and set , where denotes the set of all neighbours of in . Define to be the set of all selected vertices . It is not hard to check that is a maximal independent set of and is a partition of . Since ,
[TABLE]
This inequality can complete the proof.
- *
Corollary 2.4
.([4, 15])* For every graph , we have *
**Proof. **
Apply Theorem 2.3 with .
- *
The following corollary provides an equivalent version for Theorem 2.3.
Corollary 2.5
.* Let be a graph and let and be two real functions on . If for each , , then there is a maximal independent subset of such that*
[TABLE]
**Proof. **
Apply Theorem 2.3 with replacing instead of .
- *
3 Isolated toughness and the existence of -factors
Our aim in this section is to generalize Theorem 1.4 by giving isolated toughness conditions for existence of -factors, provided that . For this purpose, we need the following lemma due to Lovász (1970).
Lemma 3.1
.([12])* Let be a graph and let and be two integer-valued functions on with . Then has a -factor, if and only if for all disjoint subsets and of ,*
[TABLE]
The following theorem provides a common generalization for both of Theorems 2 and 3 in [13].
Theorem 3.2
.* Let be a graph and let and be two nonnegative integer-valued functions on with . Let be positive real number with . Then has a -factor, if it -iso-tough, where for each vertex ,*
[TABLE]
in which \text{\varepsilon}_{0}(v)\in\{0,1\} such that \text{\varepsilon}_{0}(v)=1 if and only if and are integers with the same parity.
**Proof. **
Let and be two disjoint subsets of . In order to apply Lemma 3.1, we should prove the inequality , where . For this purpose, we may assume that for each , . By applying Corollary 2.5 with , the graph has an independent set such that
[TABLE]
Since is -iso-tough, we have . Since is integer, we must have
[TABLE]
regardless of or not. This implies that
[TABLE]
Therefore,
[TABLE]
Hence the assertion follows from Lemma 3.1.
- *
When we consider the special case , the theorem becomes simpler as the following result.
Corollary 3.3
.* Let be a graph and let and be two nonnegative integer-valued functions on with . If is -iso-tough, then it has a -factor.*
**Proof. **
Apply Theorem 3.2 with .
- *
Corollary 3.4
.* Every -iso-tough graph has an -factor, where is a nonnegative integer-valued function on .*
**Proof. **
Apply Theorem 3.2 with setting and , , and .
- *
4 Toughness, isolated toughness, and the existence of -factors
In this section, we are going to present some sufficient toughness conditions for the existence of -factors in graphs with high enough isolated toughness.
4.1 Regular factors in -tough graphs
The following theorem significantly improves the needed toughness in Theorem 1.3 in graphs with a bit higher isolated toughness.
Theorem 4.1
.* Let be a positive integer and let be a real number with . If is a -iso-tough graph and for all ,*
[TABLE]
then has a near -factor.
**Proof. **
Let and be two disjoint subsets of such that for each , . By applying a greedy coloring, one can decompose into independent vertex sets . Let . By the assumption, we must have
[TABLE]
which implies that
[TABLE]
By the assumption, we must also have
[TABLE]
Therefore,
[TABLE]
Thus the assertion follows from Corollary 2.2.
- *
Remark 4.2
. Note that when has no complete subgraphs of order and , independent sets could be chosen such that using Brooks’ Theorem [2]. This fact allows us to reduce the lower bound on to .
4.2 Graphs with toughness less than
The following theorem gives a sufficient toughness condition for the existence of -factors.
Theorem 4.3
.* Let be a real number with 0<\text{\varepsilon}\leq 1. Let be a graph and let be a positive integer-valued function on . If is f(f+1)/\text{\varepsilon}-iso-tough and for all ,*
[TABLE]
then has a near -factor.
**Proof. **
Let and be two disjoint subsets of . We may assume that for each , where . For each , define \varphi(v)=2f(v)-\text{\varepsilon} so that . By Corollary 2.5, the graph has an independent set such that
[TABLE]
For each vertex , define t(v)=f(v)(f(v)+1)/\text{\varepsilon}-1. Since is -iso-tough, we have . Since is integer, we must have
[TABLE]
which implies that
[TABLE]
On the other hand, by the assumption,
[TABLE]
Therefore,
[TABLE]
Hence the assertion follows from Corollary 2.2.
- *
Corollary 4.4
.* Let be a graph and let be a positive integer-valued function on . If is -iso-tough and for all ,*
[TABLE]
then has a near -factor.
**Proof. **
Apply Theorem 4.3 with \text{\varepsilon}=1.
- *
The isolated toughness needed in Corollary 4.4 can be improved by a coefficient for graphs with higher toughness as the next theorem, provided that is sufficiently large.
Theorem 4.5
.* Let be a graph and let be a positive integer-valued function on , and let be a positive real number with . If is -iso-tough and for all ,*
[TABLE]
then has a near -factor.
**Proof. **
Let and be two disjoint subsets of . We may assume that for each , where . For each , define so that . By Corollary 2.5, the graph has an independent set such that
[TABLE]
For each vertex , define . Since is -iso-tough, we have . In addition, we must have
[TABLE]
which implies that
[TABLE]
On the other hand, by the assumption,
[TABLE]
Therefore,
[TABLE]
Hence the assertion follows from Corollary 2.2.
- *
4.3 Applications to the existence of connected -factors
The following lemma is a useful tool for extending factors to connected factors by inserting a matching.
Lemma 4.6
.([7], see [9])* Let be a real number with 0<\text{\varepsilon}\leq 2. Let be a simple graph and let be a factor of with minimum degree at least 2/\text{\varepsilon}+1. If for all ,*
[TABLE]
then has a connected factor containing such that for each vertex , , and for an arbitrary given vertex .
The following result shows an application of Lemma 4.6 and Theorem 4.1.
Theorem 4.7
.* Every -tough -iso-tough graph has a connected -factor, where .*
**Proof. **
We may assume that simple, by deleting multiple edges from (if necessary). By Theorem 4.1, the graph has a near -factor so that for all vertices , , except for at most one vertex with . By applying Lemma 4.6 with \text{\varepsilon}=1, the graph has a connected factor such that for each vertex , , and also . This implies that is a connected -factor.
- *
The next result shows an application of Lemma 4.6 and Corollary 4.4.
Theorem 4.8
.* Let be a real number with 0<\text{\varepsilon}\leq 2. Let be graph and let be a positive integer-valued function on with f\geq 2/\text{\varepsilon}+1. If is -iso-tough and for all ,*
[TABLE]
then has a connected -factor.
**Proof. **
We may assume that simple, by deleting multiple edges from (if necessary). By Corollary 4.4, the graph has a near -factor so that for all vertices , , except for at most one vertex with . By applying Lemma 4.6, the graph has a connected factor such that for each vertex , , and also . This implies that is a connected -factor.
- *
Corollary 4.9
.* Every -tough -iso-tough graph has a connected -factor, where is a positive integer-valued function on with .*
**Proof. **
Apply Theorem 4.8 with \text{\varepsilon}=1.
- *
5 Graphs with higher toughness
Our in this section is to provide another improvement for Theorem 1.3. Before doing so, let us refine Theorem 4.5 slightly for graphs with higher toughness.
Theorem 5.1
.* Let be a graph, let be a positive integer-valued function on , and let be a real number with . For each vertex , let \text{\varepsilon}_{0}(v)\in\{0,1\} such that \text{\varepsilon}_{0}(v)=1 if only if and are integers with the same parity. Then has a near -factor, if is \frac{1}{4(a-1)}((f(v)+a-1)^{2}-\text{\varepsilon}_{0}(v))-iso-tough and for all ,*
[TABLE]
where is the set of center vertices of the star components of in which for stars with one edge the vertex is center whenever .
**Proof. **
The proof presented here is inspired by the proof of Theorem 1 in [10]. Let and be two disjoint subsets of . For notational simplicity, we write for . Define . By Corollary 2.5, the graph has an independent set such that
[TABLE]
For each vertex , let t(v)=\frac{1}{4(a-1)}((f(v)+a-1)^{2}-\text{\varepsilon}_{0}(v)). Since is -iso-tough, we have . Since is integer, we must have
[TABLE]
which implies that
[TABLE]
Therefore, Relations (1) and (2) can deduce that
[TABLE]
Let be a maximal independent set in containing the vertices of so that . Denote by the number of components of such that for each . For every such a component , select a vertex with . Define to be the set of all selected vertices. Also, denote by the number of components of such that for each , and for at least one vertex . Set . According to this definition, it is not difficult to show that
[TABLE]
and
[TABLE]
On the other hand, by the assumption,
[TABLE]
which implies that
[TABLE]
Since for each , we must have
[TABLE]
Therefore, Relations (3) and (4) can conclude that
[TABLE]
Hence the assertion follows from Corollary 2.2.
- *
The following corollary is an improved version of Theorem 1.3.
Corollary 5.2
.* A graph has a near -factor, if for all , , and*
[TABLE]
where denotes the number of star components of .
**Proof. **
Apply Theorem 5.1 with setting when . For the special case , one can apply Theorem 1.1 directly.
- *
Corollary 5.3
.([8])* Every -tough graph of order at least has a near -factor.*
**Proof. **
We may assume that . Let be a subset of . If , then and also . Since , by the assumption, one can conclude that each vertex of contains at least neighbours. Hence . If , then we have and . Now, it is enough to apply Corollary 5.2.
- *
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1[1] J. Akiyama and M. Kano, Factors and factorizations of graphs, Springer, Heidelberg, 2011.
- 2[2] R.L. Brooks, On colouring the nodes of a network, Proc. Cambridge Philos. Soc. 37 (1941) 194–197.
- 3[3] M.-c. Cai, Connected [ k , k + 1 ] 𝑘 𝑘 1 [k,k+1] -factors of graphs, Discrete Math. 169 (1997) 1–16.
- 4[4] Y. Caro, New results on the independence number, Technical Report, Tel-Aviv University (1979).
- 5[5] V. Chvátal, Tough graphs and Hamiltonian circuits, Discrete Math. 5 (1973) 215–228.
- 6[6] M.N. Ellingham, Y. Nam, and H.-J. Voss, Connected ( g , f ) 𝑔 𝑓 (g,f) -factors, J. Graph Theory 39 (2002) 62–75.
- 7[7] M.N. Ellingham and X. Zha, Toughness, trees, and walks, J. Graph Theory 33 (2000) 125–137.
- 8[8] H. Enomoto, B. Jackson, P. Katerinis, and A. Saito, Toughness and the existence of k 𝑘 k -factors, J. Graph Theory 9 (1985) 87–95.
