746
Views
0
CrossRef citations to date
0
Altmetric
Articles

Twin g-noncommuting graph of a finite group

ORCID Icon, , , &
Pages 287-295 | Received 21 Jul 2022, Accepted 04 Nov 2022, Published online: 28 Nov 2022

Abstract

In this paper, we introduce the twin g-noncommuting graph of a finite group that is developed by combining the concepts of the g-noncommuting graph and the twin noncommuting graph of a finite group. The twin g-noncommuting graph of a finite group G, denoted by Γ¨Gg, is constructed by considering the twin vertex set as one vertex and the adjacency of the two vertices are determined from their adjacency on the g-noncommuting graph. Furthermore, we choose dihedral group, whose representation of the twin g-noncommuting graph is determined. In addition, we determine the clique number of the twin g-noncommuting graph of dihedral group.

MATHEMATICS SUBJECT CLASSIFICATION:

1. Introduction

Combining graph theory with abstract algebra is an interesting topic of study. One of the studies in abstract algebra is a group theory. A finite group can be represented as a graph by considering the group elements as vertices and the adjacency of two vertices is determined from the operation on the group. There are many researches that associated a graph and a finite group, such as the noncommuting graph by Abdollahi et al. [Citation1], the power graph by Cameron and Ghosh [Citation2], the conjugate graph by Erfanian and Tolue [Citation5], the coprime graph by Ma et al. [Citation9], and the noncentralizer graph by Tolue [Citation13]. The research on graph that represented of certain group such as the identity graph of a cyclic group by Yalcin and Kirgil [Citation16], the noncommuting graph of dihedral group by Khasraw et al. [Citation8], and the coprime graph of generalized quaternion group by Zahidah et al. [Citation17].

The concept of the noncommuting graph of a finite group is interesting to study in detail. In 1975, Paul Erd o¨ s had first introduced a graph associated to a group that is denoted by ΓG, before this concept was developed by Moghaddamfar et al. [Citation10] in terms of the noncommuting graph. Tolue et al. [Citation15] generalized the noncommuting graph to the g-noncommuting graph, denoted by ΓGg, which is a graph with the vertex set G and two distinct vertices x and y are adjacent if and only if [x,y]g and [x,y]g1.

In graph theory, two vertices a and b in a connected graph Γ are called twins if a and b have the same neighbors in V(Γ)\{a,b} [Citation12]. From that definition, Tolue [Citation14] investigated the twin vertices of the noncommuting graph and yielded the concept of the twin noncommuting graph of a finite group. The twin noncommuting of a finite group is constructed by considering the twin vertex set as one vertex and the adjacency of the two vertices are determined from their adjacency on the noncommuting graph. Moreover, Tolue [Citation14] also discussed the clique number of the twin noncommuting graph of a finite group. Based on these results, it is interesting to combine the concepts of the g-noncommuting graph and the twin noncommuting graph of a finite group that later is called by the twin g-noncommuting graph of a finite group. Moreover, in this paper we construct and determine the clique number of the twin g-noncommuting graph of dihedral group.

Throughout the paper, graphs are simple, undirected, and without loops. All of the notations and terminologies about graphs can be found in [Citation3, Citation13], and for the groups in [Citation4, Citation6, Citation7].

2. Twin g-noncommuting graph of a finite group

In this part, we introduce some definitions related to the twin g-noncommuting graph of a finite group. Let G be a finite group with the identity element e and ΓGg be the g-noncommuting graph of a group G for fixed element gG\{e}. Let x,yG, note that [x,y]=x1y1xy is the commutator of x and y of G and K(G)={[x,y]:x,yG} [Citation7]. Let Γ be a graph and xV(Γ), Tolue in [Citation14] defined N(x)={yV(Γ):d(x,y)=1} is the vertex set that is adjacent to x. A vertex uV(Γ) is called a dominant vertex if d(u,v)=1 for any other vertices vV(Γ).

Definition 2.1.

Let G be a finite group and xG. The set of elements of the group G whose commutator with x is g or g1, denoted by Lg(x), is defined as Lg(x)={yG:[x,y]=g or [x,y]=g1} where g is a non-identity element of G.

Let G be a finite group with the identity element e and ΓGg be the g-noncommuting graph of G for gG. According to [Citation15], ΓGe is not a connected graph, hence in this paper we only discuss about ΓGg for gG\{e}=G*. Meanwhile, based on Definition 2.1 we know that on ΓGg, the set Lg(x){x} where xV(ΓGg) is the vertex set that is not adjacent to x. Consequently, the vertex set that is adjacent to x of ΓGg for gG* is N(x)=G\(Lg(x){x}).

Definition 2.2.

Let G be a finite group and ΓGg be the g-noncommuting graph of G for gG*. Let aV(ΓGg), we denote the twin vertex set of a on ΓGg as a¯¨={bG:Lg(a){a,b}=Lg(b){a,b}}.

The twin vertex set on the g-noncommuting graph in Definition 2.2 is used to bring out the concept of the twin g-noncommuting graph as follows.

Definition 2.3.

Let G be a finite group, ΓGg be the g-noncommuting graph of G for gG*, and x¯¨ be the twin vertex set of x on ΓGg. The twin g-noncommuting graph of G for gG*, denoted by Γ¨Gg, is a graph with the vertex set V(Γ¨Gg)={x¯¨|xV(ΓGg} and two distinct vertices x¯¨ and y¯¨ are adjacent if and only if xyE(ΓGg).

If G is a finite abelian group, then ΓGg is a complete graph [Citation11]. Consequently, if G is a finite abelian group, then Γ¨Gg is a trivial graph. Moreover, ΓGg is a regular graph if and only if gK(G) [Citation11], so we get a corollary as follows.

Corollary 2.1.

Let G be a finite non-abelian group with the identity element e. The twin g-noncommuting graph of G for gG* is a trivial graph if and only if gK(G).

Let G be a finite non-abelian group with the identity element e. Obviously èGgèGg1 and eK(G). Furthermore, in the following results we consider èGg for a finite non-abelian group G and gK(G)\{e}=K*(G).

Lemma 2.1.

If Γ¨Gg is a twin g-noncommuting graph of a non-abelian group G for gK*(G), then e¯¨ is a dominant vertex on Γ¨Gg.

Proof.

For any x¯¨V(Γ¨Gg) and certain e¯¨V(Γ¨Gg) where e¯¨x¯¨ implies [x,e]=eg,g1. It means e¯¨ is adjacent to any x¯¨ on Γ¨Gg.

Proposition 2.1.

If Γ¨Gg is a twin g-noncommuting graph of a non-abelian group G for gK*(G), then Γ¨Gg is a connected graph with diameter two.

Proof.

Let x¯¨ and y¯¨ be two distinct vertices on Γ¨Gg. Then the following two cases occur.

  1. If x¯¨ and y¯¨ are adjacent on Γ¨Gg, then d(x¯¨,y¯¨)=1.

  2. If x¯¨ and y¯¨ are not adjacent on Γ¨Gg, then based on Lemma 2.1, x¯¨ and y¯¨ are adjacent to e¯¨V(Γ¨Gg) respectively, so d(x¯¨,y¯¨)=d(x¯¨,e¯¨)+d(y¯¨,e¯¨)=2.

Since the distance among the vertices either one or two, then the diameter of Γ¨Gg is two.

Lemma 2.2.

If Γ¨Gg is a twin g-noncommuting graph of a non-abelian group G for gK*(G), then Γ¨Gg is not a complete graph.

Proof.

Suppose Γ¨Gg is a complete graph Kn of order n, then every pair of distinct vertices x¯¨ and y¯¨ are adjacent on Γ¨Gg or in other words [x,y]g and g1. It means for all x¯¨,y¯¨V(Γ¨Gg) implies [x,y]K(G), which is a contradiction.

Proposition 2.2.

If Γ¨Gg is a twin g-noncommuting graph of a non-abelian group G for gK*(G), then Γ¨Gg is not a cycle graph.

Proof.

There are two cases, i.e.

  1. Based on Lemma 2.2, it is clear that Γ¨GgK3=C3.

  2. Suppose Γ¨GgCn where n > 3. Note that for all xZ(G) and yG\Z(G) we get [x,y]=e implies deg(x¯¨)=n1>2 on Γ¨Gg. This is contrary to the fact that the degree of any vertices on a cyclic graph is two.

3. The twin g-noncommuting graph of the dihedral group

The Dihedral group of order 2n, is denoted by D2n, can be represented as D2n=<a,b | an=b2=e,bab=a1> for nN,n  3, and e is the identity element of D2n [Citation12]. We can see that Z(D2n)={e} and K(D2n)=<a> where n is an odd number. If n is an even number, we have Z(D2n)={e,an2} and K(D2n)=<a2>. Before we discuss the twin g-noncommuting graph of D2n, we introduce a special graph as follows.

Definition 3.1.

Let k be a non-negative integer, Ωk be a k-regular graph, i.e a graph that each vertex has the same degree k and given two distinct vertices u and v where u,vV(Ωk). A graph Ak is defined as a graph with the vertex set V(Ak)=V(Ωk){u,v} and the edge set E(Ak)=E(Ωk){uw:wV(Ωk){v}}.

Let Ωk be a k-regular graph for a non-negative integer k and has order n. A graph Ak is a connected graph, consisting of a unique end vertex and a unique dominant vertex. Consequently, the maximal degree of vertex on Ak is n + 1. Therefore, we know that the order and the size of Ak are n + 2 and nk2+n+1 respectively. Furthermore, in this paper, D2n can be written as D2n={e,a,a2,,an1,b,ab,a2b,,an1b} and for all ap,aqD2n where p,q=1,2,,n and pq implies [ap,aq]=e. For simplicity, we denote K(D2n)\{e}=K*.

Lemma 3.1.

Let ΓD2ng be the g-noncommuting graph of D2n for gK* and define S1={ai:i=1,2,,n}V(ΓD2ng).

  1. If n is an odd number and g=a2jK(D2n) for j=1,2,,n12, then aj¯..={aj,anj} and e¯¨=S1\aj¯.. are twin vertex sets in S1.

  2. If n  6 is an even number and g=a2jK(D2n) for j=1,2,,n14, then aj¯..={aj,anj,an2j,an2+j} and e¯¨=S1\aj¯.. are twin vertex sets in S1.

  3. If n=4j and g=a2jK(D2n) for jN, then aj¯..={aj,anj} and e¯¨=S1\aj¯.. are twin vertex sets in S1.

Proof.

  1. Let n is an odd number and g=a2jK(D2n) for j=1,2,,n12.

    1. If [arb,ai]=g=a2j for all r=1,2,,n, then i = j.

    2. If [arb,ai]=g1=an2j for all r=1,2,,n, then i=nj.

Hence, we have Lg(aj)=Lg(anj)={arb:r=1,2,,n} and Lg(e)=Lg(ak)= for all k=1,2,,n1,kj,nj. According to Definition 2.2, the twin vertex sets in S1 are aj¯..={aj,anj} and e¯¨=S1\aj¯...

  • ii. Let n  6 is an even number and g=a2jK(D2n) for j=1,2,,n14.

    1. If [arb,ai]=g=a2j for all r=1,2,,n, then i = j or i=n2+j.

    2. If [arb,ai]=g1=an2j for all r=1,2,,n, then i=n2j or i=nj.

Hence, we have Lg(aj)=Lg(anj)=Lg(an2j)=Lg(an2+j)={arb:r=1,2,,n} and Lg(e)=Lg(ak)= for all k=1,2,,n1,kj,nj,n2j,n2+j. According to Definition 2.2, the twin vertex sets in S1 are aj¯..={aj,anj,an2j,an2+j} and e¯¨=S1\aj¯...

  • iii. For n=4j and g=a2jK(D2n) for jN. If [arb,ai]=g=g1=a2j for all r=1,2,,n, then i = j or i=nj. Hence, we have Lg(aj)=Lg(anj)={arb:r=1,2,,n} and Lg(e)=Lg(ak)= for all k=1,2,,n1,kj,nj. According to Definition 2.2, the twin vertex sets in S1 are aj¯..={aj,anj} and e¯¨=S1\aj¯...

Lemma 3.2.

Let ΓD2ng be the g-noncommuting graph of D2n for gK* and define S2={akb:k=1,2,,n}V(ΓD2ng).

  1. Let n is an odd number and g=a2jK(D2n) for j=1,2,,n12.

    1. If n=3j and j is an odd number, then amb¯..={amb,am+jb,am+2jb} for m=1,2,,j are twin vertex sets in S2.

    2. If n3j, then akb¯..={akb} for k=1,2,,n are twin vertex sets in S2.

  2. Let n  6 is an even number and g=a2jK(D2n) for j=1,2,,n14.

    1. If n=8j, then arb¯..={arb,ar+2jb,ar+4jb,ar+6jb} for r=1,2,,2j are twin vertex sets in S2.

    2. If n8j, then aqb¯..={aqb,an2+qb} for q=1,2,,n2 are twin vertex sets in S2.

  3. Let n=4j and g=a2jK(D2n) for jN. Then arb¯..={arb,ar+2jb} for r=1,2,,2j are twin vertex sets in S2.

Proof.

  1. Let n is an odd number and g=a2jK(D2n) for j=1,2,,n12.

    1. If [akb,ai]=g=a2j for all k=1,2,,n, then i = j.

    2. If [akb,ai]=g1=an2j for all k=1,2,,n, then i=nj.

    3. If [akb,alb]=g=a2j for k=1,2,,n and kl, then l=nj+k.

    4. If [akb,alb]=g1=an2j for k=1,2,,n and kl, then l=j+k.

    Hence, for any akbV(ΓD2ng) for k=1,2,,n implies Lg(akb)={aj,anj,aj+kb,anj+kb}.

    1. If n=3j and j is an odd number, then for any akbV(ΓD2ng) for k=1,2,,n implies Lg(akb){akb}=Lg(aj+kb){aj+kb}=Lg(a2j+kb){a2j+kb} and on another hand we know that aj+kb,a2j+kbLg(akb). According to Definition 2.2, the twin vertex sets in S2 are amb¯..={amb,am+jb,am+2jb} for m=1,2,,j.

    2. If n3j, then for all two distinct vertices akb and alb, where k,l=1,2,,n and kl, implies Lg(akb){akb,alb}Lg(alb){akb,alb}. According to Definition 2.2, the twin vertex sets in S2 are akb¯..={akb} for k=1,2,,n.

  2. Let n  6 is an even number and g=a2jK(D2n) for j=1,2,,n14.

    1. If [akb,ai]=g=a2j for all k=1,2,,n, then i = j or i=n2+j.

    2. If [akb,ai]=g1=an2j for all k=1,2,,n, then i=n2j or i=nj.

    3. If [akb,alb]=g=a2j for k=1,2,,n and kl, then l=nj+k or l=n2j+k.

    4. If [akb,alb]=g1=an2j for k=1,2,,n and kl, then l=j+k or l=n2+j+k.

    Hence, for all akbV(ΓD2ng) for k=1,2,,n implies Lg(akb)={aj,anj,an2j,an2+j,aj+kb,anj+kb,an2j+kb,an2+j+kb}.

    1. If n=8j, then for all akbV(ΓD2ng) where k=1,2,,n implies Lg(akb)=Lg(a2j+kb)=Lg(a4j+kb)=Lg(a6j+kb). According to Definition 2.2, the twin vertex sets in S2 are arb¯..={arb,ar+2jb,ar+4jb,ar+6jb} for r=1,2,,2j.

    2. If n8j, then for all akbV(ΓD2ng) where k=1,2,,n implies Lg(akb)=Lg(an2+kb). According to Definition 2.2, the twin vertex sets in S2 are aqb¯..={aqb,an2+qb} for q=1,2,,n2.

  3. If n=4j and g=a2jK(D2n) for jN.

    1. If [akb,ai]=g=g1=a2j for all k=1,2,,n, then i = j or i=nj.

    2. If [akb,alb]=g=g1=an2j for k=1,2,,n and kl, then l=nj+k or l=j+k.

Therefore, for all akbV(ΓD2ng) where k=1,2,,n implies Lg(akb)={aj,anj,aj+kb,anj+kb} such that Lg(akb)=Lg(a2j+kb). According to Definition 2.2, the twin vertex sets in S2 are arb¯..={arb,ar+2jb} for r=1,2,,2j.

Referring to Lemma 3.1 and Lemma 3.2, the construction of the twin g-noncommmuting graph of D2n for gK* is served in the following theorems.

Theorem 3.1.

Let Γ¨D2ng be the twin g-noncommuting graph of D2n for gK*. Let n be an odd number and g=a2j for j=1,2,,n12.

  1. If n=3j and j is an odd number, then Γ¨D2ng is Aj1 of order j + 2.

  2. If n3j, then Γ¨D2ng is An3 of order n + 2.

Proof.

Let n is an odd number and g=a2jK(D2n) for j=1,2,,n12.

  1. Let n=3j and j is an odd number. According to Lemma 3.1 and Lemma 3.2, the vertex set of Γ¨D2ng is V(Γ¨D2ng)={e¯¨,aj¯..,ab¯..,a2b¯..,,ajb¯..}. Then, there are three cases to investigate the adjacency of any vertices on Γ¨D2ng,

    1. Based on Lemma 2.1, e¯¨ is a dominant vertex on Γ¨D2ng.

    2. Since for all m=1,2,,j implies [amb,aj]=a2j=g, then for all m=1,2,,j,amb¯.. is not adjacent to aj¯...

    3. The adjacency of vertex in H={amb¯..:m=1,2,,j}V(Γ¨D2ng). Since [am1b,am2b]=a2(m1m2)a2j=g and [am1b,am2b]=a2(m1m2)an2j=g1 for m1,m2=1,2,,j and m1m2, then every two distinct vertices in H are adjacent. Consequently, the induced subgraph by H on Γ¨D2ng is Ωj1.

Based on three cases above and Definition 3.1, Γ¨D2ng is Aj1 of order j + 2.

  • ii. Let n3j, According to Lemma 3.1 and Lemma 3.2, the vertex set of Γ¨D2ng is V(Γ¨D2ng)={e¯¨,aj¯..,b¯¨,ab¯..,a2b¯..,,an1¯b..}. Then, there are three cases to investigate the adjacency of any vertices on Γ¨D2ng,

    1. Based on Lemma 2.1, e¯¨ is a dominant vertex on Γ¨D2ng.

    2. Since for all k=1,2,,n implies [akb,aj]=a2j=g, then for all k=1,2,,n,akb¯.. is not adjacent to aj¯...

    3. The adjacency of vertex in H={akb¯..:k=1,2,,n}V(Γ¨D2ng). According to Lemma 3.2, akb¯.. is adjacent to any vertices in H, except aj+kb¯.. and anj+kb¯... Consequently, the induced subgraph by H on Γ¨D2ng is Ωn3.

Based on three cases above and Definition 3.1, Γ¨D2ng is An3 of order n + 2.▪

Corollary 3.1.

Let Γ¨D2ng be the twin g-noncommuting graph of D2n for gK*.

  1. If n=3j and j is an odd number, then the induced subgraph by S={amb¯..:m=1,2,,j}V(Γ¨D2(3j)2j) is a complete subgraph of Γ¨D2(3j)2j for an odd natural number j.

  2. If n is an odd non-prime number and g=amK(D2n) for m{1,2}, then the induced subgraph by S={alb¯..,al+mb¯..,al+2mb¯..,,al+(n¯32)mb¯..} where l=1,2,,n is a complete subgraph of Γ¨D2nam.

  3. If n5 is a prime number and g=a2jK(D2n) for j=1,2,,n12, then the induced subgraph by S={a2j+lb¯..,a4j+lb¯..,a6j+lb¯..,,a(n¯1)j+lb¯..} where l=1,2,,n is a complete subgraph of Γ¨D2n2j.

Example 1.

Some constructions of the twin g-noncommuting graph of D18 for gK* can be seen in .

Figure 1. Graph (a) Γ¨D18a2, (b) Γ¨D18a4, (c) Γ¨D18a6, (d) Γ¨D18a8.

Figure 1. Graph (a) Γ¨D18a2, (b) Γ¨D18a4, (c) Γ¨D18a6, (d) Γ¨D18a8.

Theorem 3.2.

Let Γ¨D2ng be the twin g-noncommuting graph of D2n for gK*.

  1. If n=4j and g=a2jK(D2n) for jN, then Γ¨D2ng is A2j2 of order 2j+2.

  2. If n=6j and g=a2jK(D2n) for jN, then Γ¨D2ng is A3j3 of order 3j+2.

  3. If n=8j and g=a2jK(D2n) for jN, then Γ¨D2ng is A2j2 of order 2j+2.

  4. If n=2p and g=a2jK(D2n) for jN, where p  5 is a prime number and j=1,2,,p12, then Γ¨D2ng is Ap3 of order p + 2.

Proof.

  1. Let n=4j and g=a2jK(D2n) for jN. According to Lemma 3.1 and Lemma 3.2, the vertex set of Γ¨D2ng is V(Γ¨D2ng)={e¯¨,aj¯..,ab¯..,a2b¯..,,a2jb¯..}. Then, there are three cases to investigate the adjacency of any vertices on Γ¨D2ng,

    1. Based on Lemma 2.1, e¯¨ is a dominant vertex on Γ¨D2ng.

    2. Since for all r=1,2,,2j implies [arb,aj]=a2j=g, then for all r=1,2,,2j,arb¯.. is not adjacent to aj¯...

    3. The adjacency of vertices in H={arb¯..:r=1,2,,2j}V(Γ¨D2ng). According to Lemma 3.2, for all k=1,2,,n implies Lg(akb)={aj,a3j,ak+jb,ak+3jb} and ak+jb¯..={ak+jb,ak+3jb}. Therefore, arb¯.. for r=1,2,,2j is adjacent to any vertices in H, except ar+jb¯... Consequently, the induced subgraph by H on Γ¨D2ng is Ω2j2.

Based on three cases above and Definition 3.1, if n=4j and g=a2jK(D2n) for jN, then Γ¨D2ng is A2j2 of order j + 2.

  • ii. Let n=6j and g=a2jK(D2n) for jN. According to Lemma 3.1 and Lemma 3.2, the vertex set of Γ¨D2ng is V(Γ¨D2ng)={e¯¨,aj¯..,ab¯..,a2b¯..,,a3jb¯..}. Then, there are three cases to investigate the adjacency of any vertices on Γ¨D2ng,

    1. Based on Lemma 2.1, e¯¨ is a dominant vertex on Γ¨D2ng.

    2. Since for all k=1,2,,3j implies [akb,aj]=a2j=g, then for all k=1,2,,3j,akb¯.. is not adjacent to aj¯...

    3. The adjacency of vertices in H={akb¯..:k=1,2,,3j}V(Γ¨D2ng). According to Lemma 3.2, for all l=1,2,,n implies Lg(alb)={aj,a2j,a4j,a5j,aj+lb,ak+3jb,a2j+l,a4j+l} and aj+lb¯..={aj+lb,a4j+l},a5j+lb¯..={a2j+lb,a5j+lb}. Therefore, akb¯.. for k=1,2,,3j is adjacent to any vertices in H, except aj+kb¯.. and a5j+kb¯... Consequently, the induced subgraph by H on Γ¨D2ng is Ω3j3.

Based on the above three cases and Definition 3.1, if n=6j and g=a2jK(D2n) for jN, then Γ¨D2ng is A3j3 of order 3j+2.

  • iii. Let n=8j and g=a2jK(D2n) for jN. According to Lemma 3.1 and Lemma 3.2, the vertex set of Γ¨D2ng is V(Γ¨D2ng)={e¯¨,aj¯..,ab¯..,a2b¯..,,a2jb¯..}. Then, there are three cases to investigate the adjacency of any vertices on Γ¨D2ng,

    1. Based on Lemma 2.1, e¯¨ is a dominant vertex on Γ¨D2ng.

    2. Since for all r=1,2,,2j implies [arb,aj]=a2j=g, then for all r=1,2,,2j,arb¯.. is not adjacent to aj¯...

    3. The adjacency of vertices in H={arb¯..:r=1,2,,2j}V(Γ¨D2ng). According to Lemma 3.2, for all k=1,2,,n implies Lg(akb)={aj,a3j,a5j,a7j,aj+kb,a7j+kb,a3j+k,a5j+k} and aj+kb¯..={aj+kb,a3j+kb,a5j+kb,a7j+kb}. Therefore, arb¯.. for r=1,2,,2j is adjacent to any vertices in H, except ar+jb¯... Consequently, the induced subgraph by H on Γ¨D2ng is Ω2j2.

Based on three cases above and Definition 3.1, if n=8j and g=a2jK(D2n) for jN, then Γ¨D2ng is A2j2 of order 2j+2.

  • iv. Let n=2p and g=a2jK(D2n) for jN, where p  5 is prime number and j=1,2,,p12. According to Lemma 3.1 and Lemma 3.2, the vertex set of Γ¨D2ng is V(Γ¨D2ng)={e¯¨,aj¯..,ab¯..,a2b¯..,,apb¯..}. Then, there are three cases to investigate the adjacency of any vertices on Γ¨D2ng,

    1. Based on Lemma 2.1, e¯¨ is a dominant vertex on Γ¨D2ng.

    2. Since for all k=1,2,,p implies [akb,aj]=a2j=g, then for all k=1,2,,p,akb¯.. is not adjacent to aj¯...

    3. The adjacency of vertices in H={akb¯..:k=1,2,,p}V(Γ¨D2ng). According to Lemma 3.2, for all l=1,2,,n implies Lg(alb)={aj,anj,apj,ap+j,aj+lb,anj+lb,apj+lb,ap+j+lb},

aj+lb¯..={aj+lb,ap+j+lb}, and apj+lb¯..={anj+lb,apj+lb}. Therefore, akb¯.. for k=1,2,,p is adjacent to any vertices in H, except ak+jb¯.. and ap+(kj)b¯... Consequently, the induced subgraph by H on Γ¨D2ng is Ωp3.

Based on three cases above and Definition 3.1, if n=2p and g=a2jK(D2n) for jN, where p  5 is a prime number and j=1,2,,p12, then Γ¨D2ng is Ap3 of order p + 2.▪

Corollary 3.2.

If n = mj and g=a2jK(D2n) where jN and m={4,6}, then the induced subgraph by {aj+lb¯..,aj+1+lb¯..,,a2j1+lb¯..} where l=1,2,,mj2 is a complete subgraph of Γ¨D2n2j.

Corollary 3.3.

Let Γ¨D2ng be the twin g-noncommuting graph of D2n for g=K*. Then

  1. èD2(4j)a2jèD2(8j)a2j for jN

  2. If n  5 is a prime number, then Γ¨D2na2jΓ¨D2(2n)a2j for j=1,2,,n12.

4. The clique number of the twin g-noncommuting graph of the dihedral group

In this part, we discuss related to the clique number of the twin g-noncommuting graph of dihedral group. Referring to Definition 3.1 we have that graph Ak=(ΩkK1)+K1, where K1 is a trivial graph. Hence, we get clique number of Ak as follows.

Corollary 4.1.

Let Ωk be a k-regular graph, then ω(Ak)=ω(Ωk)+1.

Henceforth, we observe the clique number of the twin g-noncommuting graph of dihedral group regarding to the previous section. Note that notation Γ¨D2ng claims the twin g-noncommuting graph of D2n for gK(D2n)\{e}=K*. A vertex aj¯..V(Γ¨D2ng) is an end vertex, so it is not possible to be a vertex candidate for clique on Γ¨D2ng. Meanwhile a vertex e¯¨V(Γ¨D2ng) is a dominant vertex, so e¯¨ is a vertex candidate for clique on Γ¨D2ng. Based on these two conditions, the largest clique’s proof in this section only observe to the adjacency of the twin vertex sets in S2={akb:k=1,2,,n} in Lemma 3.2.

Theorem 4.1.

If n=3j and g=a2jK(D2n) for jN and j is an odd number, then ω(Γ¨D2(3j)a2j)=j+1.

Proof.

According to Corollary 3.1, the induced subgraph by S={amb¯..:m=1,2,,j}V(Γ¨D2(3j)a2j) is a complete subgraph. It means the spanning vertex set of a clique Δ on Γ¨D2(3j)a2j is V(Δ)=S{e¯¨} where |V(Δ)|=j+1. In another hand, we know that |V(Γ¨D2(3j)a2j)|=j+2 and based on Lemma 2.2, Γ¨D2(3j)a2j is not a complete graph. Consequently, a clique Δ is the largest clique on Γ¨D2(3j)a2j (see Γ¨D18a6 as an evidence of this theorem for j = 3).▪

Theorem 4.2.

If n is an odd non-prime number and g=a,a2K(D2n), then ω(Γ¨D2ng)=n+12.

Proof.

Let n is an odd non-prime number.

  1. Let g=aK(D2n) and g1=an1K(D2n). According to Corollary 3.1, the spanning vertex set of a clique Δ on Γ¨D2ng is V(Δ)={e¯¨}H where H={alb¯..,al+1b¯..,,al+n32b¯..} and l=1,2,,n. Consequently, |V(Δ)|=n+12 and vertices on H are sequential.

Hereafter, suppose there is another clique Λ where V(Λ)V(Γ¨D2ng) and V(Λ)>V(Δ). Without loss of generality, let V(Λ)=V(Δ){amb¯..} for m{l+n32+r:r=1,2,,n+12}, then there are two cases for m,

  1. If m{l+n32+s:s=1,2,,n12}, then amb¯..V(Λ) is not adjacent to am(n12)b¯..V(Λ) since [amb,amn12b]=an1=g1 which is a contradiction.

  2. If m=l+n1, then amb¯..V(Λ) is not adjacent to al+(n32)b¯..V(Λ) since [amb,al+n32b]=[al+n1b,al+n32b]=a=g which is a contradiction.

Based on two cases above, Λ is not a clique of Γ¨D2ng. Hence, clique Δ is the largest clique of Γ¨D2ng, thus ω(Γ¨D2ng)=n+12.

  • ii. Let g=a2K(D2n) and g1=an2K(D2n). According to Corollary 3.1, the spanning vertex set of a clique Δ on Γ¨D2ng is V(Δ)={e¯¨}H where H={alb¯..,al+2b¯..,al+4b¯..,,al+n¯3b¯..} and l=1,2,,n. Consequently, |V(Δ)|=n+12 and vertices on H are sequential.

Hereafter, suppose there is another clique Λ where V(Λ)V(Γ¨D2ng) and V(Λ)>V(Δ). Without loss of generality, let V(Λ)=V(Δ){amb¯..} for m{l+(2r1):r=1,2,,n12}{l+n1}, then there are two cases for m,

  1. If m{l+(2r1):r=1,2,,n12}, then amb¯..V(Λ) is not adjacent to am+1b¯..V(Λ) since [amb,am+1b]=an2=g1 which is a contradiction.

  2. If m=l+n1, then amb¯..V(Λ) is not adjacent to alb¯..V(Λ) since [amb,alb]=[al+n1b,alb]=an2=g1 which is a contradiction.

Based on two cases above, Λ is not a clique of Γ¨D2ng. Hence, clique Δ is the largest clique of Γ¨D2ng, thus ω(Γ¨D2ng)=n+12.

Theorem 4.3.

If n  5 is a prime number and g=a2jK(D2n) for j=1,2,,n12, then ω(Γ¨D2ng)=n+12.

Proof.

According to Corollary 3.1, the spanning vertex set of a clique Δ on Γ¨D2ng is V(Δ)={e¯¨}H where H={a2j+lb¯..,a4j+lb¯..,a6j+lb¯..,,a(n1)j+lb¯..} and l=1,2,,n. Consequently, |V(Δ)|=n+12 and vertices on H are sequential.

Hereafter, suppose there is another clique Λ where V(Λ)V(Γ¨D2ng) and V(Λ)>V(Δ). Without loss of generality, let V(Λ)=V(Δ){amb¯..} for m{r+l:r=j,3j,5j,,nj}, then there are two cases for m,

  1. If m{r+l:r=j,3j,5j,,(n2)j}, then amb¯..V(Λ) is not adjacent to am+jb¯..V(Λ) since [amb,am+jb]=an2j=g1 which is a contradiction.

  2. If m=nj+l, then amb¯..V(Λ) is not adjacent to a(n1)j+lb¯..V(Λ) since [amb,a(n1)j+lb]=[anj+lb,a(n1)j+lb]=a2j=g1 which is a contradiction.

Based on two cases above, Λ is not a clique of Γ¨D2ng. Hence, clique Δ is the largest clique of Γ¨D2ng, thus ω(Γ¨D2ng)=n+12.

Example 2.

All possibilities of the largest clique of twin g-noncommuting graph of dihedral group for n = 7 can be seen in .

Figure 2. All possibilities of the largest clique of Γ¨D14a6.

Figure 2. All possibilities of the largest clique of Γ¨D14a6.

Theorem 4.4.

If n=4j and g=a2jK(D2n) for jN, then ω(Γ¨D2(4j)a2j)=j+1.

Proof.

Let S={arb¯..:r=1,2,,2j}V(Γ¨D2(4j)a2j), then there are two cases,

  1. If j = 1, then Γ¨D8a2j is A0 i.e, a star graph. Thus, obviously the clique number of Γ¨D8a2j is 2.

  2. If j > 1, then according to Corollary 3.2, the spanning vertex set of a clique Δ on Γ¨D2(4j)a2j is V(Δ)={e¯¨}H where H={aj+lb¯..,aj+1+lb¯..,,a2j1+lb¯..} and l=1,2,,2j. Consequently, |V(Δ)|=j+1 and vertices on H are sequential.

Hereafter, suppose there is another clique Λ where V(Λ)V(Γ¨D2(4j)a2j) and V(Λ)>V(Δ). Without loss of generality, let V(Λ)=V(Δ){amb¯..} for m{r+l:r=1,2,,j1}{2j+l}. Then there are two cases for m,

  • case 1. If m{r+l:r=1,2,,j1}, then amb¯..V(Λ) is not adjacent to am+jb¯..V(Λ) since [amb,aj+mb]=a2j=g1 which is a contradiction.

  • case 2. If m=2j+l, then amb¯..V(Λ) is not adjacent to aj+lb¯..V(Λ) since [amb,aj+lb]=[a2j+lb,aj+lb]=a2j=g1 which is a contradiction.

Based on two cases above, Λ is not a clique of Γ¨D2(4j)a2j. Hence, clique Δ is the largest clique of Γ¨D2(4j)a2j, thus ω(Γ¨D2(4j)a2j)=j+1.

Corollary 4.2.

If n=8j and g=a2jK(D2n) for jN, then ω(Γ¨D2(8j)a2j)=j+1.

Theorem 4.5.

If n=6j and g=a2jK(D2n) for jN, then ω(Γ¨D2(6j)a2j)=j+1.

Proof.

According to Corollary 3.2, then the spanning vertex set of a clique Δ on Γ¨D2(6j)a2j is V(Δ)={e¯¨}H where H={aj+lb¯..,aj+1+lb¯..,,a2j1+lb¯..} and l=1,2,,3j. Consequently, |V(Δ)|=j+1 and vertices on H are sequential.

Hereafter, suppose there is another clique Λ where V(Λ)V(Γ¨D2(6j)a2j) and V(Λ)>V(Δ). Without loss of generality, let V(Λ)=V(Δ){amb¯..} for m{r+l:r=0,1,2,,j1}{s+l:s=2j,2j+1,,3j1}. Then there are two cases for m,

  1. If m{r+l:r=0,1,2,,j1}, then amb¯..V(Λ) is not adjacent to am+jb¯..V(Λ) since [am+jb,amb]=a2j=g which is a contradiction.

  2. If m{s+l:s=2j,2j+1,,3j1}, then amb¯..V(Λ) is not adjacent to amjb¯..V(Λ) since [amjb,amb]=an2j=g1 which is a contradiction.

Based on two cases above, Λ is not a clique of Γ¨D2(6j)a2j. Hence, clique Δ is the largest clique of Γ¨D2(6j)a2j, thus ω(Γ¨D2(6j)a2j)=j+1.

Corollary 4.3.

If n=3j and g=a2jK(D2n) for jN, then ω(Γ¨D2(3j)a2j)=j+1.

5. Conclusion

In this research, we have built up the new concept of twin g-noncommuting graph of a group. We also have constructed the twin g-noncommuting graph of the dihedral group. Furthermore, we have determined the clique number of the twin g-noncommuting graph of the dihedral group.

Disclosure statement

The authors report there are no competing interests to declare.

Data availability statement

Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.

References

  • Abdollahi, A., Akbari, S, Maimani, H. R. (2006). Noncommuting graph of a group. J. Algebra 298(2): 468–492.
  • Cameron, P. J, Ghosh, S. (2011). The power graph of a finite group. Discrete Math. 311(13): 1220–1222.
  • Chartrand, G., Lesniak, L, Zhang, P. (2016). Graphs and Digraphs. 6th ed. London: Chapman and Hall/CRC.
  • Dummit, D. S, Foote, R. M. (2004). Abstract Algebra. 3rd ed. Hoboken, NJ: John Wiley and Sons Inc.
  • Erfanian, A, Tolue, B. (2012). Conjugate graphs of finite groups. Discrete Math. Algorithms Appl. 4(2): 1250035.
  • Fraleigh, J. B. (2014). A First Course in Abstract Algebra. 7th ed. Upper Saddle River, NJ: Pearson Education Limited.
  • Guralnick, R. M. (1982). Commutators and commutator subgroups. Adv. Math. (3)45: 319–330. 1982)
  • Khasraw, S. M. S., Ali, I. D, Haji, R. R. (2020). On the noncommuting graph of dihedral group. Electr. J. Graph Theory Appl. 8(2): 233–239.
  • Ma, X., Wei, H, Yang, L. (2014). The coprime graph of a group. Int. J. Group Theory 3(3): 13–23.
  • Moghaddamfar, A. R., Shi, W. J., Zhou, W, Zokayi, A. R. (2005). On the noncommuting graph associated with a finite group. Sib. Math. J. 46(2): 325–332.
  • Neumann, B. H. (1976). A problem of Paul Erdos on groups. J. Aust. Math. Soc. 21(4): 467–472.
  • Okamoto, F., Crosse, L., Phinezy, B., Zhang, P. Kalamazoo. (2010). The local metric dimension of a graph. Math. Bohemica 135(3): 239–255.
  • Tolue, B. (2015). The noncentralizer graph of a finite group. Math. Rep. 17(67): 265–275.
  • Tolue, B. (2020). The twin noncommuting graph of a group. Rend. Circ. Mat. Palermo, II. Ser. 69(2): 591–599.
  • Tolue, B., Erfanian, A, Jafarzadeh, A. (2014). A kind of noncommuting graph of finite groups. J. Sci. Islamic Republic of Iran 25(4): 379–384.
  • Yalcin, N. F, Kirgil, Y. (2020). Identity graphs of finite cyclic groups. J. Balikesir Inst. Sci. Technol. 22(1): 149–158.
  • Zahidah, S., Mahanani, D. M, Oktaviana, K. L. (2021). Connectivity indices of coprime graph of generalized quaternion group. J. Indonesian Math. Soc. 27(3): 285–296.