
Citation: T. Deepa, Raúl M. Falcón, M. Venkatachalam. On the r-dynamic coloring of the direct product of a path with either a complete graph or a wheel graph[J]. AIMS Mathematics, 2021, 6(2): 1470-1496. doi: 10.3934/math.2021090
[1] | G. Nandini, M. Venkatachalam, Raúl M. Falcón . On the r-dynamic coloring of subdivision-edge coronas of a path. AIMS Mathematics, 2020, 5(5): 4546-4562. doi: 10.3934/math.2020292 |
[2] | Meiqin Jin, Ping Chen, Shuangliang Tian . Interval edge colorings of the generalized lexicographic product of some graphs. AIMS Mathematics, 2024, 9(11): 30597-30611. doi: 10.3934/math.20241477 |
[3] | T. Deepa, M. Venkatachalam, Raúl M. Falcón . On the r-dynamic coloring of the direct product of a path with either a path or a cycle. AIMS Mathematics, 2020, 5(6): 6496-6520. doi: 10.3934/math.2020419 |
[4] | Shuangliang Tian, Ping Chen . Edge-coloring of generalized lexicographic product of graphs. AIMS Mathematics, 2024, 9(6): 15988-15995. doi: 10.3934/math.2024774 |
[5] | Syed Ahtsham Ul Haq Bokhary, Zill-e-Shams, Abdul Ghaffar, Kottakkaran Sooppy Nisar . On the metric basis in wheels with consecutive missing spokes. AIMS Mathematics, 2020, 5(6): 6221-6232. doi: 10.3934/math.2020400 |
[6] | Gaixiang Cai, Fengru Xiao, Guidong Yu . The identification numbers of lollipop graphs. AIMS Mathematics, 2025, 10(4): 7813-7827. doi: 10.3934/math.2025358 |
[7] | Kiki A. Sugeng, Fery Firmansah, Wildan, Bevina D. Handari, Nora Hariadi, Muhammad Imran . Several properties of antiadjacency matrices of directed graphs. AIMS Mathematics, 2024, 9(10): 27834-27847. doi: 10.3934/math.20241351 |
[8] | Yanyi Li, Lily Chen . Injective edge coloring of generalized Petersen graphs. AIMS Mathematics, 2021, 6(8): 7929-7943. doi: 10.3934/math.2021460 |
[9] | Shahbaz Ali, Muhammad Khalid Mahmmod, Raúl M. Falcón . A paradigmatic approach to investigate restricted hyper totient graphs. AIMS Mathematics, 2021, 6(4): 3761-3771. doi: 10.3934/math.2021223 |
[10] | Hongyu Chen, Li Zhang . A smaller upper bound for the list injective chromatic number of planar graphs. AIMS Mathematics, 2025, 10(1): 289-310. doi: 10.3934/math.2025014 |
In 2001, Bruce Montgomery [1] (see also [2]) introduced the concept of r-dynamic proper k-coloring of a graph G=(V(G),E(G)) as any proper k-coloring c:V(G)→{0,…,k−1} of such a graph such that
|c(N(v))|≥min{r,d(v)}, | (1.1) |
for all v∈V(G). Here, N(v) and d(v) denote, respectively, the neighborhood and the degree of the vertex v. In addition, he introduced the notion of r-dynamic chromatic number χr(G) of the graph G as the minimum positive integer k for which an r-dynamic proper k-coloring of G exists. As such, these concepts constitute a natural generalization of the classical notions of proper coloring and the chromatic number χ(G) of a graph G, which arise when r=1. Montgomery himself proposed the natural question about how much the difference between χr(G) and χ(G) varies, for every r>1, and also whether such a difference is bounded for all graphs. Concerning this last question, he proved [1] the existence of graphs for which this difference is unbounded even for r=2. Further, Montgomery also introduced the study of the r-dynamic chromatic number of specific families of graphs, for all r>1. More specifically, he determined explicitly all these values for any complete graph, cycle or tree [1] and dealt with the case r=2 for any multipartite graph [2].
Since the original manuscript of Montgomery, a wide amount of graph theorists have dealt with the study of upper bounds concerning the r-dynamic chromatic number of any graph. In this regard, Montgomery himself [1] proved that χ2(G)≤Δ(G)+3, for every graph G, where Δ(G) denotes the maximum vertex degree in G. Shortly after, Lai et al. [2] proved that χ2(G)≤Δ(G)+1, if Δ(G)≥4. In addition, Montgomery [1] also conjectured that χ2(G)≤χ(G)+2, for every regular graph G. This inequality was proved by Lai et al. [3] for graphs that are connected and claw-free. Concerning the conjecture itself, it was proved by Akabari et al. [4] in case of dealing with bipartite regular graphs. Furthermore, Alishahi [5] proved the existence of a constant c such that χ2(G)≤χ(G)+c⋅ln(k)+1, for every k-regular graph G. This last author [6] also proved that χ2(G)≤χ(G)+γ(G), where γ(G) denotes the domination number of G. For a general positive integer r, Jahanbekam et al. [7] proved that χr(G)≤r⋅Δ(G)+1. For r≥2, Lai et al. [3] had already proved that χr(G)≤Δ(G)+r2−r+1, whenever Δ(G)≤r. Finally, concerning the study of upper bounds of the r-dynamic chromatic number of products of graphs, Akbari et al [8] proved that χ2(G◻H)≤max{χ2(G),χ2(H)}, if δ(G)≥2, where G◻H denotes the Cartesian product of two graphs G and H, and δ(G) denotes the minimum vertex degree of the graph G.
In the recent literature, it is also remarkable the acquired relevance of determining explicitly the r-dynamic chromatic number of specific families of graphs, for every positive integer r. It is so that this value has already been studied for grid graphs [9,7]; helm graphs [10]; prism graphs, three-cyclical ladder graphs, joint graphs and circulant graphs [11]; toroidal graphs [12]; some cycle-related graphs [13]; coronations of paths [14]; and subdivision-edge coronas of a path [15]. In addition, it has also been studied the r-dynamic chromatic number of the Cartesian product and the corona product of distinct types of graphs [16,17,18,19,20,21,22]. Nevertheless, the r-dynamic chromatic number of direct products of graphs has only recently been given attention. More specifically, it has explicitly been determined [23] for the direct product of any given path with either a path or a cycle. This paper delves into this topic by determining the r-dynamic chromatic number of the direct product of any given path with either a complete graph or a wheel graph.
The paper is organized as follows. In Section 2, we describe some preliminary concepts and results on Graph Theory that are used throughout the paper. Then, Sections 3 and 4 deal, respectively, with the r-dynamic chromatic number of the direct product of a path with either a complete graph or a wheel graph.
This section deals with some preliminary concepts and results on Graph Theory that are used throughout the paper. For more details about this topic, we refer the reader to the manuscripts [24,25].
A graph is any pair G=(V(G),E(G)) formed by a set V(G) of vertices and a set E(G) of edges so that each edge joins two vertices, which are then said to be adjacent. From now on, let vw be the edge formed by two vertices v,w∈V(G). If v=w, then the edge constitutes a loop. A graph is called simple if it does not contain loops. Further, the number of vertices of a graph is its order. A graph is called finite if its order is finite. This paper deals with the direct product G×H of two finite and simple graphs G=(V(G),E(G)) and H=(V(H),E(H)). Its vertex set is the Cartesian product V(G)×V(H). Two vertices (v,v′) and (w,w′) in such a set are adjacent if and only if vw∈E(G) and v′w′∈E(H). Figure 1 illustrates this last concept.
The set of vertices that are adjacent to a vertex v∈V(G) constitutes its neighborhood NG(v). The cardinality dG(v) of this set is the degree of the vertex v. If there is no risk of confusion, then we use the respective notations N(v) and d(v). Furthermore, we denote, respectively, δ(G) and Δ(G) the minimum and maximum vertex degree of the graph G. The following result follows straightforwardly from the previous definitions.
Lemma 1. Let G and H be two finite simple graphs. Then,
a) dG×H((v,w))=dG(v)dH(w), for all (v,w)∈V(G×H).
b) δ(G×H)=δ(G)δ(H).
c) Δ(G×H)=Δ(G)Δ(H).
A finite graph is called complete if all its vertices are pairwise adjacent. A path between two distinct vertices v and w of a given graph G is any ordered sequence of adjacent and pairwise distinct vertices ⟨v0=v,v1,…,vn−2,vn−1=w⟩ in V(G), with n>2. If v=w, then such a sequence is called a cycle. A graph is said to be connected if there always exists a path between any pair of vertices. Further, if all the vertices of a cycle are joined to a new vertex, then the resulting graph is called a wheel. Such a new vertex is called the center of the wheel graph. From here on, let Kn, Pn, Cn and Wn respectively denote the complete graph, the path, the cycle and the wheel graph of order n.
A proper k-coloring of a graph G is any map c:V(G)→{0,…,k−1} assigning k colors to the set of vertices V(G) so that no two adjacent vertices have identical color. The minimum positive integer k for which such a proper k-coloring exists is the chromatic number χ(G) of the graph G. Particular cases of proper coloring and chromatic number are the so-called r-dynamic proper k-coloring and the r-dynamic chromatic number, which have already been described in the introductory section (see (1.1)). The following results are known.
Lemma 2. [1] Let G be a graph and let r be a positive integer. Then,
min{r,Δ(G)}+1≤χr(G)≤χr+1(G). |
Moreover,
χr(G)≤χΔ(G)(G). |
Lemma 3. [3] Let n and r be two positive integers. Then, the following results hold.
a) If n>2, then χr(Pn)={2,ifr=1,3,otherwise.
b) χr(Kn)=n.
Lemma 4. [18] Let n>2 be a positive integer. Then,
χ2(Wn)={3,ifnis odd,4,otherwise. |
In case of dealing with the r-dynamic chromatic number of a direct product of graphs, the following results hold.
Lemma 5. [23] Let G and H be two finite simple graphs and let r be a positive integer such that r≤δ(G′), for some G′∈{G,H}. Then,
χr(G×H)≤χr(G′). |
Theorem 6. [23] Let m, n and r be three positive integers such that m,n>2. Then,
χr(Pm×Cn)={2,ifr=1,3,ifr=2andn=3t,for somet≥1,4,if{r=2andn≠3t,for allt≥1,r=3andn∉{4,5,10},5,if{r=3andn∈{4,5,10},for allt≥1,r≥4andn=5t,for somet≥1,r≥4,m∈{3,4}andn∉{3,4,6,7,8,14},6,if{r≥4,m∈{3,4}andn∈{3,4,6,7,8,14},r≥4,m≥5and n≠5t,for allt≥1. |
In this section, we study the r-dynamic chromatic number of the direct product of a path
Pm=⟨u0,…,um−1⟩ |
and a complete graph Kn of set of vertices
V(Kn)={v0,…,vn−1}, |
where m and n are two positive integers such that m>2. From Lemma 1, we have that
δ(Pm×Kn)=n−1≤2(n−1)=Δ(Pm×Kn). |
More specifically, for each vertex (ui,vj)∈Pm×Kn, it is
d((ui,vj))={n−1, if i∈{0,m−1},2(n−1), if 0<i<m−1. |
Firstly, we focus on the case n≤3, which follows readily from already known results.
Theorem 7. Let m and r be two positive integers such that m>2. Then, the following assertions hold.
a) χr(Pm×K1)=1.
b) χr(Pm×K2)={2,ifr=1,3,otherwise.
c) χr(Pm×K3)={r+1,ifr≤3,6,otherwise.
Proof. The first assertion follows simply from the fact that the direct product Pm×K1 is not connected. Further, the second assertion follows from Lemma 3 once it is observed that the direct product Pm×K2 is formed by a pair of disjoint paths of order m. Finally, the third assertion follows from Theorem 6 once it is observed that the complete graph K3 constitutes a cycle of order three.
Now, let us prove a pair of preliminary lemmas that are useful to deal with the case n>3. In order to simplify the notation, for each given proper coloring c of the direct product Pm×Kn, we denote from now on ci,j:=c(ui,vj), for all non-negative integers i<m and j<n.
Lemma 8. Let m, n and r be three positive integers such that m>2, n>3 and 1<r≤n−2. Then,
r+2≤χr(Pm×Kn). |
Proof. Since r≤n−2<Δ(Pm×Kn), we have from Lemma 2 that r+1≤χr(Pm×Kn). Then, in order to prove the result, let us suppose the existence of an r-dynamic proper (r+1)-coloring c of the direct product Pm×Kn. Since |c(N((u0,v0)))|=r, we can suppose, without loss of generality, that c1,j=j, for all positive integer j≤r. As a consequence, since c is a proper coloring, it must be c0,0=0. Moreover, there must exist a positive integer j0≤r such that c1,r+1=j0. Now, since |c(N((u1,vr+1)))|=r>1 and c is a proper coloring, we have that j∈{c0,j,c2,j}⊆{0,j}, for all j∈{1,…,r}∖{j0}, and c0,j0=c2,j0=c0,r+1=c2,r+1=0. But then, again from the fact that c is a proper coloring, it must be c1,j=j0, for all j∈{0,r+2,…,n−1}. It implies that |c(N((u0,vj)))|=r−1<r, for all positive integer j≤r such that j≠j0, which contradicts Condition (1.1).
Lemma 9. Let m, n and r be three positive integers such that m>2, n>3 and ⌊3n2⌋≤r≤2n−2. Then,
r+2≤χr(Pm×Kn). |
Proof. Since r≤2n−2=Δ(Pm×Kn), we have from Lemma 2 that r+1≤χr(Pm×Kn). Let us suppose the existence of an r-dynamic proper (r+1)-coloring c of the direct product Pm×Kn. Similarly to the proof of Lemma 8, we can suppose, without loss of generality, that c1,j=j, for all non-negative integer j<n, and c0,j=n+j, for all non-negative integer j<r−n. Then, since |c(N((u1,vn−1)))∖{n,…,r−1}|=n and the map c is a proper coloring, it must be c2,j=j, for all non-negative integer j<r−n, and j∈{c0,j,c2,j}, for all j∈{r−n,…,n−1}. But then, since n+j must be a color in the set c(N((u1,vj))), for all non-negative integer j<r−n, it should also be n+j∈{c0,r−n,…,c0,n−1,c2,r−n,…,c2,n−1}, for all non-negative integer j<r−n. It is only possible if and only if r−n+1≤2n−r−1. That is, it should be r≤3n2−1, which contradicts the hypothesis.
The following result establishes the r-dynamic chromatic number of the direct product of a path and a complete graph of order n>3.
Theorem 10. Let m, n and r be three positive integers such that m>2 and n>3. Then,
χr(Pm×Kn)={r+1,if eitherr=1,orn−1≤r<⌊3n2⌋,r+2,if either1<r≤n−2,or⌊3n2⌋≤r≤2n−2,2n,otherwise. |
Proof. Let us study separately each case by defining an appropriate r-dynamic proper coloring c of the corresponding direct product Pm×Kn satisfying Condition (1.1).
● Case r=1.
This case follows simply from Lemmas 2 and 5 once we notice that 1≤δ(Pm) and 2=χ(Pm).
● Case 1<r≤n−2.
From Lemma 8, we have that r+2≤χr(Pm×Kn). Then, let the map c be defined so that
ci,j={j, if 1≤j≤r,0, if i is even and j∉{1,…,r},r+1, otherwise. |
Condition (1.1) holds and hence, χr(Pm×Kn)=r+2. Figure 2 illustrates the direct product P3×K5, for r=3.
● Case n−1≤r<⌊3n2⌋.
From Lemma 2, we have that r+1≤χr(Pm×Kn). Then, let the map c be defined recursively as follows.
- For each non-negative integer j<n, we have that c0,j=j.
- For each positive integer i<m and each non-negative integer j<n, we have that
ci,j={(n+(i−1)(r−n+1)+k)mod(r+1), if {j=((i−1)(r−n+1)+k)modn,j= for some 0≤k≤r−n,ci−1,j otherwise. |
Condition (1.1) holds and hence, χr(Pm×Kn)=r+1. Figures 3–5 illustrate the direct product P6×K5, for r∈{4,5,6}.
● Case ⌊3n2⌋≤r≤2n−2.
From Lemma 9, it is r+2≤χr(Pm×Kn). Then, let the map c be defined so that, for each pair of non-negative integers i<m and j<n, we have that
ci,j={j, if either imod4∈{0,1}, or j>r−n+1,n+j, otherwise. |
Condition (1.1) holds and hence, χr(Pm×Kn)=r+2. Figure 6 illustrates the direct product P5×K7, for r=10.
● Case r>2n−2.
The result follows from the previous case and Lemma 2, once we notice that Δ(Pm×Kn)=2n−2.
Let m and n be two positive integers such that m>2 and n>3. In this section, we study the r-dynamic chromatic number of the direct product of a path
Pm=⟨u0,…,um−1⟩ |
and a wheel graph Wn of set of vertices
V(Wn)={v0,…,vn−2,v}. |
Here, v denotes the center of the wheel graph. Thus, Wn contains the cycle graph
Cn−1=⟨v0,…,vn−2⟩. |
From Lemma 1, we have that
δ(Pm×Wn)=3≤2(n−1)=Δ(Pm×Wn). |
More specifically, for each vertex (ui,vj)∈Pm×Wn, it is
d((ui,vj))={3, if i∈{0,m−1},6, otherwise. |
In addition, for each vertex (ui,v)∈Pm×Wn, it is
d((ui,v))={n−1, if i∈{0,m−1},2(n−1), otherwise. |
Firstly, let us prove a result that enables us to focus on those direct products Pm×Wn such that either n=7, or n is even, or n=4t+1, for some t≥1.
Lemma 11. Let m>2, n>3 and r be three positive integers such that n is odd. Then,
χr(Pm×W2n+1)=χr(Pm×Wn+1). |
Proof. The result follows straightforwardly from the fact that the direct product Pm×W2n+1 may be considered as two direct products Pm×Wn+1, whose sets of vertices are disjoint except for those vertices corresponding to the common center of their respective wheel graphs.
The following preliminary lemmas establish certain bounds for the r-dynamic chromatic number χr(Pm×Wn). In order to simplify the notation, for each given proper coloring c of the direct product Pm×Wn, we denote from now on ci,j:=c(ui,vj), for all non-negative integers i<m and j<n−1. In addition, all the indices of the vertices vj associated to the wheel graph Wn are considered to be taken modulo n−1. Let us start with a pair of bounds of the r-dynamic chromatic number χr(Pm×Wn) arising from the fact that every wheel graph contains a cycle.
Lemma 12. Let m, n and r be three positive integers such that m>2, n>3 and r>1. Then,
χr(Pm×Wn)≤χr−1(Pm×Cn−1)+1, |
whenever there exists an (r−1)-dynamic proper χr−1(Pm×Cn−1)-coloring c of the direct product Pm×Cn−1 such that the following two conditions hold.
a) min{r,n−1}≤|{ci,j:0≤j<n−1}|, for all i∈{1,m−2}.
b) min{r,2(n−1)}≤|{ci−1,j,ci+1,j:0≤j<n−1}|, for all positive integer i≤m−2.
Proof. Let us suppose the existence of the map c in the hypothesis. Then, let ¯c:V(Pm×Wn)→{0,…,χr−1(Pm×Cn−1)} be defined so that, for each non-negative integer i<m, we have that ¯c((ui,v))=χr−1(Pm×Cn−1), and ¯ci,j=ci,j, for all non-negative integer j<n−1. It is simply verified that this map ¯c is a proper coloring of the direct product Pm×Wn satisfying Condition (1.1), for every vertex (ui,vj)∈V(Pm×Wn). Moreover, the compliance of both conditions (a) and (b) enable us to ensure that Condition (1.1) also holds for every vertex (ui,v)∈V(Pm×Wn), and hence, the result holds. Figure 7 illustrates this constructive proof for the direct products P4×C5 and P4×W6.
Let us establish now a series of lower bounds of χr(Pm×Wn), for r≥2.
Lemma 13. Let m>2 and n>3 be two positive integers such that n is even. Then,
3<χ2(Pm×Wn). |
Proof. From Lemma 2, we have that 3≤χr(Pm×Wn). Let us suppose the existence of a 2-dynamic proper 3-coloring c of the direct product Pm×Wn. Since n is even, there must exist a non-negative integer j<n−1 such that c1,j≠c1,j+2 (otherwise, |c(N(u0,v))|=1, which contradicts Condition (1.1)). Then, without loss of generality, we can take c1,0=0, c1,2=1 and c((u0,v))=2. As a consequence, c0,1=c((u2,v))=2≠c((u1,v)). Again, without loss of generality, we can suppose that c((u1,v))=0. Under such assumptions, if there exists a non-negative integer j0<n−1 such that c0,j0=1, then it should be c1,j0−1=c1,j0+1=0. But then, |c(N((u0,vj0)))|=1, which is a contradiction. Hence, c0,j=2, for all non-negative integer j<n−1. In a similar way, we may ensure that c2,j=2, for all non-negative integer j<n−1. But then, |c(N((u1,vj)))|=1, for all non-negative integer j<n−1, which constitutes again a contradiction. As a consequence, no such a map c exists and hence, the result holds.
Lemma 14. Let m>2 and n>3 be two positive integers such that n≠3k+1, for all k>0. Then,
4<χ3(Pm×Wn). |
Proof. From Lemma 2, we have that 4≤χr(Pm×Wn). Let us suppose the existence of a 3-dynamic proper 4-coloring c of the direct product Pm×Wn. Condition (1.1), together with the fact that c is a proper coloring, implies that ci,j≠c((u1,v)), for all non-negative integers i<3 and j<n−1. As a consequence, for each non-negative integer j<n−1, we have that c1,j≠c0,j+1=c2,j+1≠c1,j+2≠c1,j. Thus, since |c(N((u1,vj+2)))|=3, it must be c0,j+3=c2,j+3=c1,j and c1,j+4=c0,j+1. In short, c1,j, c1,j+2 and c1,j+4 are pairwise distinct, for all non-negative integer j<n−1. This contradicts the fact of being n≠3k+1, for all k>0, and hence, the result holds.
Lemma 15. Let m>2 be a positive integer and let n∈{5,6,11}. Then,
5<χ4(Pm×Wn). |
Proof. From Lemma 2, we have that 5≤χ4(Pm×Wn). So, let us suppose the existence of a 4-dynamic proper 5-coloring c of the direct product Pm×Wn. Since d((u0,vj))=3, for all non-negative integer j<n−1, it must be c((u1,v))∉{c1,j:0≤j<n−1}. As a consequence, since |c(N((u0,v)))|=4, we have that c((u0,v))=c((u1,v))=c((u2,v)). The following study of cases arise.
● Case n=5.
In a recursive way, it is simply verified that |{ci,j:0≤j<4}|=4 and c((ui,v))=c((u0,v)), for all non-negative integer i<m. Thus, the map ¯c:V(Pm×C4)→{0,1,2,3,4}∖{c((u0,v))} that is defined so that ¯ci,j=ci,j, for all non-negative integers i<m and j<4 is a 4-dynamic proper 4-coloring of the direct product Pm×C4. It contradicts Theorem 6 and hence, the result holds.
● Case n=6.
Since |c(N((u0,v)))|=4, we have that {c1,j:0≤j≤4}={0,1,2,3}. Thus, there must exist a pair of non-negative integers j1,j2≤4 such that c1,j1=c1,j2. Since |c(N((u0,vj)))|=3, for all non-negative integer j<5, it must be (j2−j1)mod5∈{1,4}. Without loss of generality, let us suppose that that c1,0=c1,4. Then, since c is a proper coloring and ci,j≠c((u1,v)), for all (i,j)∈{0,2}×{0,1,2,3,4}, we have that {c0,1,c2,1,c0,3,c2,3}⊆{c1,1,c1,3}. It contradicts the fact of being |c(N((u0,v)))|=4 and hence, the result holds.
● Case n=11.
It follows simply from the case n=6 and Lemma 11.
Lemma 16. Let m>2 and n>4 be two positive integers such that one of the following assertions hold.
a) 5≤n≠5t+1, for all t≥1.
b) m∈{3,4} and n∈{5,7,8,9,15}.
Then,
6<χ5(Pm×Wn). |
Proof. Lemma 2 implies that 6≤χ5(Pm×Wn). So, let us suppose the existence of a 5-dynamic proper 6-coloring c of the direct product Pm×Wn.
If n=5, then Condition (1.1) implies that |c(N((u0,v)))|=4 and |c(N((u1,v0)))|=5. The first equality enables us to ensure that c1,j≠c1,k, for all non-negative integers j,k<4. But then, the second equality implies that c1,2∈N((u1,v0)), which is not possible because N((u1,v0))=N((u1,v2)) and c is a proper coloring. As a consequence, it must be 6<χ5(Pm×W5).
In n>5, then Condition (1.1) implies that |c(N((u0,v)))|=5 and hence, since c is a proper coloring, we have that c(u2,v)=c(u0,v). Let j0<n−1 be a non-negative integer. Since |c(N((u1,vj0)))|=5, we have that c(N(u1,vj0))={0,1,2,3,4,5}∖{c1,j0,c(u0,v)}. Thus,
{ci,j:i∈{0,2},0≤j<n−1}={0,1,2,3,4,5}∖{c(u0,v)}. |
As a consequence, c(u1,v)=c(u0,v). Moreover, if m>3, then c2,j0−1≠c2,j0+1. Thus, if m∈{3,4} (respectively m≥5) the map ¯c:V(P3×Cn−1)→{0,1,2,3,4}∖{c((u0,v))} that is defined so that ¯ci,j=ci,j, for all non-negative integers i<4 and j<n−1, would be a 4-dynamic proper 5-coloring of the direct product P3×Cn (respectively, P4×Cn). It contradicts Theorem 6 when n∈{5,7,8,9,15} (respectively, n≠5t+1, for all t≥1) and hence, the result holds.
Lemma 17. Let m>2, n≥4 and r≥6 be three positive integers such that r≤2(n−1). Then,
r+1<χr(Pm×Wn). |
Proof. Lemma 2 implies that r+1≤χr(Pm×Wn). So, let us suppose the existence of an r-dynamic proper (r+1)-coloring c of the direct product Pm×Wn. Notice that c(u0,v)≠c(u2,v). Otherwise, we would have, for instance, that |c(N(u1,v0))|<r, which contradicts Condition (1.1). Now, if c(u1,v)≠c(u0,v), then Condition (1.1) implies that c(u0,v)∈c(N(u1,v)). But then, |c(N(u1,vj))|<r, for some non-negative integer j<n−1, which contradicts the mentioned condition. Hence, c(u1,v)=c(u0,v). Then, again from Condition (1.1) we would have that c(u2,v)∈c(N(u1,v)) and hence, |c(N(u1,vj))|<r, for some non-negative integer j<n−1. It contradicts once more time Condition (1.1). As a consequence, no r-dynamic proper (r+1)-coloring exists.
Further, in order to establish an upper bound based on Lemma 5, the following proposition determines the 3-dynamic chromatic number of a wheel graph.
Proposition 18. Let m and n be two positive integers such that m>2 and n>3. Then,
χ3(Pm×Wn)≤χ3(Wn)={4,ifn=3k+1,for somek>0,5,if6≠n≠3k+1,for allk>0,6,ifn=6. |
Proof. Since δ(Wn)=3, for all n>3, Lemma 5 implies that χ3(Pm×Wn)≤χ3(Wn). In addition, from Lemma 2, we have that 4≤χ3(Wn). Let us study separately each case by describing to this end an appropriate 3-dynamic proper coloring c of the wheel graph Wn satisfying Condition (1.1). An illustrative example of each case is shown in Figure 8.
● Case n=3k+1, for some k>0.
Let the map c be defined so that c(v)=3 and c(vj)=jmod3, for all non-negative integer j<n−1. Condition (1.1) holds and hence, χ3(Wn)=4. Figure 8 illustrates the case n=7.
● Case n≠3k+1, for all k>0.
From Condition (1.1), we have that c(v), c(vj), c(vj+1) and c(vj+2) are pairwise distinct, for all non-negative integer j<n−1. As a consequence, since n≠3k+1, for all k>0, it must be 4<χ3(Wn). The following subcases arise.
- Subcase n=3k+2, for some k>0.
Let the map c be defined so that c(v)=4 and
c(vj)={jmod3, if j≠n−2,3, otherwise. |
Condition (1.1) holds and hence, χ3(Pm×Wn)=5. Figure 8 illustrates the case n=8.
- Subcase n=3k, for some k>2.
Let the map c be defined so that c(v)=4 and
c(vj)={jmod3, if j<n−6,j−n+5, if n−5≤j<n−1,3, if j=n−6. |
Condition (1.1) holds and hence, χ3(Pm×Wn)=5. Figure 8 illustrates the case n=9.
- Subcase n=6.
The result holds because, from Condition (1.1), no two vertices in Wn can share the same color in any given 3-dynamic proper coloring of the wheel graph W6. An illustrative example is shown in Figure 8.
After enumerating all the previous bounds, we are in condition of determining exactly the r-dynamic chromatic number of the direct product Pm×Wn. Firstly, let us establish the r-dynamic chromatic number of the direct product Pm×W4, which follows readily from Theorem 10, once we notice that the wheel graph W4 coincides with the complete graph K4.
Theorem 19. Let m and r be two positive integers such that m>2. Then,
χr(Pm×W4)={r+1,ifr∈{1,3,4,5},r+2,ifr∈{2,6},8,otherwise. |
The following theorem is the main result of the section. It establishes the r-dynamic chromatic number of the direct product of a path and a wheel graph of order n≠4.
Theorem 20. Let m, n and r be three positive integers such that m>2 and n>4. Then,
χr(Pm×Wn)={2,ifr=1,3,ifr=2andnis odd,4,if{r=2andnis even,r=3andn=3k+1,for somek>0,5,if{r=3andn≠3k+1,for allk>0,r=4andn∉{5,6,11},6,if{r=4andn∈{5,6,11},r=5and{n=5t+1,for somet≥1,m∈{3,4}andn∉{5,7,8,9,15},7,if{r=5and{n∈{5,7,8,9,15},m≥5andn≠5t+1,for allt≥1,r+2,if6≤r<2(n−1),2n,otherwise. |
Proof. Let us study separately each case by defining to this end an appropriate r-dynamic proper coloring c of the direct product Pm×Wn satisfying Condition (1.1).
● Case r=1.
This case follows simply from Lemmas 2 and 5 once we notice that 1≤δ(Pm) and 2=χ(Pm).
● Case r=2.
From Lemma 2, we have that 3≤χ2(Pm×Wn). Moreover, if n is even, then Lemma 13 implies that 4≤χ2(Pm×Wn). The result holds because, since r=2<3=δ(Wn), we have from Lemmas 4 and 5 that
χ2(Pm×Wn)≤{3, if n is odd,4, if n is even. |
● Case r=3.
From Lemma 2, we have that 4≤χ3(Pm×Wn). Hence, from Proposition 18, we have that χ3(Pm×W3k+1)=4, for all positive integer k. Further, Lemma 14 implies that 5≤χ3(Pm×Wn), whenever n≠3k+1, for all k>0. Hence, again from Proposition 18, we have that χ3(Pm×Wn)=5, for all positive integer n distinct from 6 and 3k+1, for all k>0.
Finally, since the 2-dynamic proper 4-coloring of the direct product Pm×C5 that is described in the proof of Theorem 17 in [23] satisfies both conditions (a) and (b) of Lemma 12, this last result enables us to ensure that χ3(Pm×W6)≤5 and hence, that this upper bound is reached.
● Case r=4.
From Lemma 2, we have that 5≤χ4(Pm×Wn). Then, the following study of cases arise.
- Subcase n≠3k+1, for all k>0, except for n∈{5,6,11}.
Again from the constructive proof of Theorem 17 in [23] and Lemma 12, we can ensure that χ4(Pm×Wn)≤5. From Lemma 2, this upper bound is reached.
- Subcase n=6k+1, for some k>0.
Let the map c be defined so that c((ui,v))=4, for all non-negative integer i<m, and such that the following assertions hold.
* ci,j=ci,jmod6, for all non-negative integers i<m and j<n.
* c1,3=c2,3=0.
* c0,3=c2,4=c3,4=1.
* c0,1=c0,4=c1,4=c3,5=c4,5=2.
* c0,2=c1,2=c0,5=c1,5=c2,5=3.
* Let j<6, l<6 and t<⌊m6⌋ be three non-negative integers such that 6t+j+l<m. Then,
c6t+j+l,j={(j−2t)mod4, if l<4,(j−2t−1)mod4, otherwise. |
Condition (1.1) holds and hence, χ4(Pm×Wn)=5. Figure 9 illustrates the direct product P7×W7.
- Subcase n=6k+4, for some k>0.
Let the map c be defined so that, for each non-negative integer i<m, we have that c((ui,v))=4, and
ci,(j+i)mod(n−1)={imod4, if j=0,(i+3)mod4, if j=6k, for some k>0,(i+2)mod4, if j=6k+2, for some k≥0,(i+1)mod4, if j=6k+4, for some k≥0. |
Condition (1.1) holds, and hence, χ4(Pm×Wn)=5. Figure 10 illustrates the direct product P17×W16.
- Subcase n∈{5,6,11}.
From Lemma 15, we have that 6≤χ4(Pm×Wn). In order to prove the case n=5, let the map c be defined so that the following assertions hold.
* ci,j=j, for all non-negative integers i<m and j<n−1.
* For each non-negative integer i<m, we have that
c(ui,v)={4, if imod4∈{0,1},5, otherwise. |
Condition (1.1) holds, and hence, χ4(Pm×W5)=6. Figure 11 illustrates the direct product P4×W5.
Further, both cases n=6 and n=11 follow from Lemma 12 and the respective 4-dynamic proper 5-colorings that are described in the proof of Theorem 17 in [23].
● Case r=5.
From Lemma 2, we have that 6≤χ5(Pm×Wn). Then, Lemmas 11 and 16 enable us to focus on the following study of cases.
- Subcase n=5t+1, for some t≥1, or m∈{3,4} and n∉{5,7,8,9,15}.
Lemma 12, together with the 4-dynamic proper 5-colorings that are described in the proof of Theorem 17 in [23], enables us to ensure that χ5(Pm×Wn)=6.
- Subcase n=5.
From Lemma 16, we have that 7≤χ5(Pm×W5). Then, let the map c be defined so that
ci,j={(2⌊i2⌋+j)mod6, if j∈{0,1},(2⌈i2⌉−1−j)mod6, if j∈{2,3},6, otherwise. |
Condition (1.1) holds, and hence, χ5(Pm×W5)=7. Figure 12 illustrates the direct product P6×W5.
- Subcase n=4t+1, for some t≥2.
From Lemma 16, we have that 7≤χ5(Pm×Wn). Let the map c be defined so that
c((ui,v))={5, if imod4∈{0,1},6, otherwise. |
In addition, for each (ui,vj)∈V(Pm×Wn), we have that
ci,j={2imod 5, if j∈{0,1},(2i+1)mod 5, if j∈{2,3},(2i+2)mod 5, if j∈{4,5},(2i+3)mod 5, if j=6 or j=8k+l, for some k≥0 and l∈{1,2},(2i+4)mod 5, if j=7 or j=8k+l, for some k≥0 and l∈{0,3}. |
Condition (1.1) holds, and hence, χ5(Pm×Wn)=7. Figures 13 illustrates the direct product P6×W13.
- Subcase n=7.
Again from Lemma 16, we have that 7≤χ5(Pm×W5). Then, let the map c be defined so that c(ui,v)=6, for all positive integer i<m, and
ci,j={(j+k)mod6, if i=2k, for some k≥0,(j+k−2)mod6, if i=2k+1, for some k≥0. |
Condition (1.1) holds, and hence, χ5(Pm×W7)=7. Figure 14 illustrates the direct product P6×W7.
- Subcase n=6t+k, for some positive integer t and k∈{0,2,4}.
Let the map c be defined so that
c((ui,v))={5, if imod4∈{0,1},6, otherwise. |
In addition, for each (ui,vj)∈V(Pm×Wn), we have that
ci,j={2imod 5, if (j,k)∈{(0,0),(0,2),(0,4),(1,2),(1,4)},(2i+1)mod 5, if (j,k)∈{(1,0),(2,2),(2,4),(3,2),(3,4)},(2i+2)mod 5, if (j,k)∈{(2,0),(4,2),(4,4),(5,4)},(2i+3)mod 5, if (j,k)∈{(3,0),(5,2),(6,4),(7,4)},(2i+4)mod 5, if (j,k)∈{(4,0),(6,2),(8,4)},(2i+l)mod 5, if j=6t+5+k+l, for some l∈{0,1,2,3,4,5}. |
Condition (1.1) holds, and hence, χ5(Pm×Wn)≤7. In particular, Lemmas 11 and 16 imply that χ5(Pm×Wn)=7, for all m≥5. In addition, they enable us to ensure that χ5(Pm×W5)=7, for all n∈{8,15} and m∈{3,4}. Figures 15–17 illustrate the direct products P6×Wn, for n∈{12,14,16}.
● Case r=6.
From Lemma 17, we have that 8≤χr(Pm×Wn). Then, Lemma 11 enables us to focus on the following study of cases. In all of them, the described map c satisfies that
c((ui,v))={6, if imod4∈{0,1},7, otherwise. |
- Subcase n=7.
Let the map c be defined so that
ci,j={(j+k)mod6, if i=2k, for some k≥0,(j+k−2)mod6, if i=2k+1, for some k≥0. |
Condition (1.1) holds, and hence, χ6(Pm×W7)=8. Figure 18 illustrates the case m=7.
- Subcase n=4t+1, with t≥2.
Let the map c be defined so that
ci,j={imod6, if j=0,(i+3)mod6, if j=1,(i−1)mod6, if j=4k+2, for some k≥0,(i+2)mod6, if j=4k+3, for some k≥0,(i−2)mod6, if j=4k for some k>0,(i+1)mod6, if j=4k+1 for some k>0. |
Condition (1.1) holds, and hence, χ6(Pm×Wn)=8. Figure 19 illustrates the case m=7.
- Subcase n=6t, with t≥1.
Let the map c be defined so that
ci,j={(j+k)mod6, if j≤2 and i=2k, for some k≥0,(j+k+1)mod6, if j∈{3,4} and i=2k, for some k≥0,(j+k−2)mod6, if j≤3 and i=2k+1, for some k≥0,(3+k)mod6, if j=4 and i=2k+1, for some k≥0. |
and, for each pair of non-negative integers j and k≤5 such that 6j+k+5<n−1, we have that
ci,6j+k+5={(k+l)mod6, if i=2l, for some l≥0,(k+l−2)mod6, if i=2l+1, for some l≥0. |
Condition (1.1) holds, and hence, χ6(Pm×Wn)=8. Figure 20 illustrates the direct product P7×W12.
- Subcase n=6t+2, with t≥1.
Let the map c be defined so that, for each non-negative integer j<n−1, we have that
c0,j=(j−1)mod6 |
and
c1,j={c0,j, if jmod6∈{1,3,4,5},5, if j≡2(mod6),1, otherwise. |
In addition, the following assertions hold.
* Let k be a positive integer such that 2k<m. Then,
c2k,j=c2k−1,(j+3)mod6. |
* Let k be a positive integer such that 2k+1<m. Then,
c2k+1,j={c2k,j, if jmod6∈{1,3,4,5},c2k,2, if j≡0(mod6),c2k,0, otherwise. |
Condition (1.1) holds, and hence, χ6(Pm×Wn)=8. Figure 21 illustrates the direct product P6×W14.
- Subcase n=6t+4, with t≥1.
Let c be the map defined in the previous subcase (n=6t+2). Then, let the map c′ be defined so that, for each pair of non-negative integers i<m and j<n−1, we have that
c′i,j={ci,j−2, if j≥2,ci,j+4, otherwise. |
Condition (1.1) holds, and hence, χ6(Pm×Wn)=8. Figure 22 illustrates the direct product P6×W16.
● Case 7≤r≤2(n−1).
From Lemma 17, we have that r+2≤χr(Pm×Wn). The following study of cases arises. In all of them, the described map c satisfies that
c((ui,v))={r, if imod4∈{0,1},r+1, otherwise. |
- Subcase r≤n−1.
Let t<r be such that n−1≡t(modr). Then, let the map c be defined so that
ci,j={k, if i=0 and j=2k+l, for some k<t and l∈{0,1},j−t, if i=0 and 2t≤j,(ci−1,j−2)modr, otherwise. |
Condition (1.1) holds, and hence, χr(Pm×Wn)=r+2. Figure 23 illustrates the direct product P5×W12, for r=8.
- Subcase n+2≤r.
Let c be the map just described in the previous subcase (r≤n−1). Then, let the map c′ be defined so that
c′i,j={ci,j+n−1, if imod4∈{1,2} and ci,j<r−n+1,ci,j, otherwise. |
Condition (1.1) holds, and hence, χr(Pm×Wn)=r+2. Figure 24 illustrates the direct product P6×W7, for r=8.
This paper has delved into the study of the r-dynamic chromatic number of the direct product of two given graphs. More specifically, it has explicitly been determined the r-dynamic chromatic number of the direct product of any given path Pm with either a complete graph Kn or a wheel Wn. In this regard, Theorems 3.4, 4.9 and 4.10 are the main results of the manuscript. Particularly, it has been obtained that r+1≤χr(Pm×G)≤2n, for all G∈{Kn,Wn}. In addition, we have also established in Proposition 18 the 3-dynamic chromatic number of any wheel graph.
Similarly to the previous work of the authors in the topic [23], a significant number of technical results is required in order to prove the main theorems of the paper. This fact enables us to corroborate that the problem of r-dynamic coloring the direct product of two given graphs is not trivial at all. Of particular interest for the continuation of this paper is the study of the r-dynamic coloring of the direct product of two complete graphs and that one concerning the direct product of two wheel graphs. The r-dynamic coloring of the direct product of a path, a complete graph or a wheel with other types of graphs is also established as related further work.
The authors want to express their gratitude to the anonymous referees for the comprehensive reading of the paper and their pertinent comments and suggestions, which helped improve the manuscript. Further, Falcón's work is partially supported by the research project FQM-016 from Junta de Andalucíi a, and the Departmental Research Budget of the Department of Applied Mathematics I of the University of Seville.
The authors declare no conflict of interest.
[1] | B. Montgomery, Dynamic coloring of graphs, ProQuest LLC, Ann Arbor, MI: Ph.D Thesis, West Virginia University, 2001. |
[2] | H.-J. Lai, B. Montgomery, H. Poon, Upper bounds of dynamic chromatic number, Ars Combin., 68 (2003), 193-201. |
[3] |
H.-J. Lai, J. Lin, B. Montgomery, T. Shui, S. Fan, Conditional colorings of graphs, Discrete Math., 306 (2006), 1997-2004. doi: 10.1016/j.disc.2006.03.052
![]() |
[4] |
S. Akbari, M. Ghanbari, S. Jahanbakam, On the dynamic chromatic number of graphs, Contemp. Math., 531 (2010), 11-18. doi: 10.1090/conm/531/10454
![]() |
[5] |
M. Alishahi, On the dynamic coloring of graphs, Discrete Appl. Math., 159 (2011), 152-156. doi: 10.1016/j.dam.2010.10.012
![]() |
[6] |
M. Alishahi, Dynamic chromatic number of regular graphs, Discrete Appl. Math., 160 (2012), 2098-2103. doi: 10.1016/j.dam.2012.05.017
![]() |
[7] |
S. Jahanbekama, J. Kim, O. Suil, D. B. West, On r-dynamic coloring of graphs, Discrete Appl. Math., 206 (2016), 65-72. doi: 10.1016/j.dam.2016.01.016
![]() |
[8] | S. Akbari, M. Ghanbari, S. Jahanbekam, On the dynamic chromatic number of Cartesian product graphs, Ars Combin., 114 (2014), 161-167. |
[9] |
R. Kang, T. Müller, D. B. West, On r-dynamic coloring of grids, Discrete Appl. Math., 186 (2015), 286-290. doi: 10.1016/j.dam.2015.01.020
![]() |
[10] | N. Mohanapriya, J. Vernold Vivin, M. Venkatachalam, δ-dynamic chromatic number of helm graph families, Cogent Mathematics & Statistics, 3 (2016), 1178411. |
[11] |
D. Dafik, D. E. W. Meganingtyas, K. D. Purnomo, M. D. Tarmidzi, I. H. Agustin, Several classes of graphs and their r-dynamic chromatic numbers, J. Phys.: Conf. Ser., 855 (2017), 012011. doi: 10.1088/1742-6596/855/1/012011
![]() |
[12] |
S. Loeb, T. Mahoney, B. Reiniger, J. Wise, Dynamic coloring parameters for graphs with given genus, Discrete Appl. Math., 235 (2018), 129-141. doi: 10.1016/j.dam.2017.09.013
![]() |
[13] |
N. Mohanapriya, J. V. Vivin, J. Kok, M. Venkatachalam, On dynamic coloring of certain cyclerelated graphs, Arabian J. Math., 9 (2020), 213-221. doi: 10.1007/s40065-018-0219-3
![]() |
[14] |
B. J. Septory, A. I. Dafik, I. H. Agustin, D. A. R. Wardani, On r-dynamic chromatic number of coronation of order two of any graphs with path graph, IOP Conf. Series: Earth and Environmental Science, 243 (2019), 012113. doi: 10.1088/1755-1315/243/1/012113
![]() |
[15] | G. Nandini, M. Venkatachalam, R. M. Falcón, On the r-dynamic coloring of subdivision-edge coronas of a path, AIMS Math., 5 (2020), 4546-4562. |
[16] | I. H. Agustin, D. Dafik, A. Y. Harsya, On r-dynamic coloring of some graph operation, Int. J. Combin., 1 (2016), 22-30. |
[17] |
I. H. Agustin, D. A. R. Wardani, B. J. Septory, A. I. Kristiana, E. Y. Kurniawati, The r-dynamic chromatic number of corona of order two of any graphs with complete graph, J. Phys.: Conf. Ser., 1306 (2019), 012046. doi: 10.1088/1742-6596/1306/1/012046
![]() |
[18] | K. Kaliraj, H. Naresh Kumar, J. Vernold Vivin, On dynamic colouring of Cartesian product of complete graph with some graphs, J. Taibah Univ. Sci., 14 (2020), 168-171. |
[19] | A. I. Kristiana, D. Dafik, M. I. Utoyo, I. H. Agustin, On r-dynamic chromatic number of the corronation of path and several graphs, Int. J. Adv. Eng. Res. Sci., 4 (2017), 237123. |
[20] | A. I. Kristiana, M. I. Utoyo, Dafik, The lower bound of the r-dynamic chromatic number of corona product by wheel graphs, AIP Conf. Proc., 2014 (2018), 020054. |
[21] |
A. I. Kristiana, M. I. Utoyo, D. Dafik, On the r-dynamic chromatic number of the corronation by complete graph, J. Phys. Conf. Ser., 1008 (2018), 012033. doi: 10.1088/1742-6596/1008/1/012033
![]() |
[22] |
A. I. Kristiana, M. I. Utoyo, R. Alfarisi, D. DAFIK, r-dynamic coloring of the corona product of graphs, Discrete Mathematics, Algorithms and Applications, 12 (2020), 2050019. doi: 10.1142/S1793830920500196
![]() |
[23] | T. Deepa, M. Venkatachalam, R. M. Falcón, On the r-dynamic coloring of the direct product of a path with either a path or a cycle, AIMS Math., 5 (2020), 6496-6520. |
[24] | J. A. Bondy, U. S. R. Murty, Graph theory with applications, New York: Macmillan Ltd. Press, 1976. |
[25] | F. Harary, Graph Theory, Reading, Massachusetts: Addison Wesley, 1969. |
1. | Raúl M. Falcón, M. Venkatachalam, S. Gowri, G. Nandini, On the r-dynamic coloring of some fan graph families, 2021, 29, 1844-0835, 151, 10.2478/auom-2021-0039 | |
2. | T. Deepa, M. Venkatachalam, Ismail Naci Cangul, On r-dynamic chromatic number of some brick product graphs C(2n, 1, p), 2022, 15, 1793-5571, 10.1142/S1793557122501029 | |
3. | Raúl M. Falcón, M. Venkatachalam, S. Gowri, G. Nandini, On the r-dynamic coloring of the direct product of a path and a k-subdivision of a star graph, 2022, 14, 1793-8309, 10.1142/S1793830921501214 | |
4. | Raúl M. Falcón, Venkitachalam Aparna, Nagaraj Mohanapriya, Optimal secret share distribution in degree splitting communication networks, 2023, 18, 1556-1801, 1713, 10.3934/nhm.2023075 | |
5. | Raúl M. Falcón, S. Gowri, M. Venkatachalam, Solving the dynamic coloring problem for direct products of paths with fan graphs, 2023, 31, 1844-0835, 115, 10.2478/auom-2023-0006 |