Loading [MathJax]/jax/output/SVG/jax.js
Research article

Characterization of Q graph by the burning number

  • Received: 22 November 2023 Revised: 25 December 2023 Accepted: 26 December 2023 Published: 16 January 2024
  • MSC : 05C45, 05C70, 05C75

  • The burning number b(G) of a graph G, introduced by Bonato, is the minimum number of steps to burn the graph, which is a model for the spread of influence in social networks. In 2016, Bonato et al. studied the burning number of paths and cycles, and based on these results, they proposed a conjecture on the upper bound for the burning number. In this paper, we determine the exact value of the burning number of Q graphs and confirm this conjecture for Q graph. Following this, we characterize the single tail and double tails Q graph in term of their burning number, respectively.

    Citation: Yinkui Li, Jiaqing Wu, Xiaoxiao Qin, Liqun Wei. Characterization of Q graph by the burning number[J]. AIMS Mathematics, 2024, 9(2): 4281-4293. doi: 10.3934/math.2024211

    Related Papers:

    [1] Kai An Sim, Kok Bin Wong . On the cooling number of the generalized Petersen graphs. AIMS Mathematics, 2024, 9(12): 36351-36370. doi: 10.3934/math.20241724
    [2] Sufyan Asif, Muhammad Khalid Mahmood, Amal S. Alali, Abdullah A. Zaagan . Structures and applications of graphs arising from congruences over moduli. AIMS Mathematics, 2024, 9(8): 21786-21798. doi: 10.3934/math.20241059
    [3] Bilal A. Rather, M. Aijaz, Fawad Ali, Nabil Mlaiki, Asad Ullah . On distance signless Laplacian eigenvalues of zero divisor graph of commutative rings. AIMS Mathematics, 2022, 7(7): 12635-12649. doi: 10.3934/math.2022699
    [4] Saba Al-Kaseasbeh, Ahmad Erfanian . A bipartite graph associated to elements and cosets of subgroups of a finite group. AIMS Mathematics, 2021, 6(10): 10395-10404. doi: 10.3934/math.2021603
    [5] Baishuai Zuo, Chuancun Yin . Stein’s lemma for truncated generalized skew-elliptical random vectors. AIMS Mathematics, 2020, 5(4): 3423-3433. doi: 10.3934/math.2020221
    [6] Mohamed R. Zeen El Deen, Ghada Elmahdy . New classes of graphs with edge δ graceful labeling. AIMS Mathematics, 2022, 7(3): 3554-3589. doi: 10.3934/math.2022197
    [7] M. Haris Mateen, Muhammad Khalid Mahmmod, Doha A. Kattan, Shahbaz Ali . A novel approach to find partitions of Zm with equal sum subsets via complete graphs. AIMS Mathematics, 2021, 6(9): 9998-10024. doi: 10.3934/math.2021581
    [8] Derya Sağlam . Minimal homothetical and translation lightlike graphs in Rn+2q. AIMS Mathematics, 2022, 7(9): 17198-17209. doi: 10.3934/math.2022946
    [9] Saud Fahad Aldosary, Ram Swroop, Jagdev Singh, Ateq Alsaadi, Kottakkaran Sooppy Nisar . An efficient hybridization scheme for time-fractional Cauchy equations with convergence analysis. AIMS Mathematics, 2023, 8(1): 1427-1454. doi: 10.3934/math.2023072
    [10] 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
  • The burning number b(G) of a graph G, introduced by Bonato, is the minimum number of steps to burn the graph, which is a model for the spread of influence in social networks. In 2016, Bonato et al. studied the burning number of paths and cycles, and based on these results, they proposed a conjecture on the upper bound for the burning number. In this paper, we determine the exact value of the burning number of Q graphs and confirm this conjecture for Q graph. Following this, we characterize the single tail and double tails Q graph in term of their burning number, respectively.



    All graphs considered in this paper are finite and simple. We use Bondy and Murty [1] for notation and terminology not defined here. The concept of burning graph is introduced by Bonato [2] as a model for social contagion, and it was also studied by Janssen and Roshanbin [3,4]. Graph burning is inspired by graph theoretic processes like graph cleaning [5], and firefighting [6]. Given a finite connected graph, the process of burning a graph begins with all vertices being unburned. At time step 1, a single vertex is chosen to be burned. In each subsequent time step, a new vertex is burned and previously burning vertices burn their neighbors. This implies that if a vertex is already burned at step t1, then its unburned neighbors (if any) become automatically burned at the end of step t. The process is completed when all vertices of the graph have been burned. The minimum number of time steps required to burn all the vertices of a graph G is called the burning number of G and denoted by b(G). The significance of the burning number is that the speed of the spread of a contagion varies in reverse to the burning number of the respective graph model. For example, b(Kn)=2, when n2. Suppose that we burn a graph G in k steps in a burning process. The sequence (x1,,xk) is called a burning sequence for G. The burning number of a graph G, denoted by b(G), is also the length of a shortest burning sequence for G. Such a burning sequence is called an optimal burning sequence for G.

    The burning problem is NP-complete even for trees and path-forests [4]. Thus, it is also interesting to determine the burning number of special classes of graphs. Bonato and Lidbetter [7] considered the bounds on the burning numbers of spiders (which are trees with exactly one vertex of degree strictly greater than two) and path-forests. Sim et al. [8] studied the burning number of generalized Petersen graphs. Mitsche et al. [9] established the burning number of graph products. Recently, Mitsche et al. [10] focused on a few probabilistic aspects of the burning problem. Very recently, Liu et al. [11,12] considered the burning number of theta graphs, path forests, and spiders.

    In 2016, Bonato et al. [3] studied the burning number of paths and cycles, and based on these results, they proposed a conjecture on the upper bound for the burning number.

    Theorem 1.1. For a path Pn or a cycle Cn of order n, we have b(Pn)=b(Cn)=n. Moreover, if G is a graph of order n with a Hamiltonian path, then b(G)n.

    Conjecture 1.2. For a connected graph G of order n, b(G)n.

    The Q graph, often denoted by Qv(s,t1,t2), is obtained by joining one vertex of cycle Cs+1 with one vertex of path Pt+1 as v, where t1+t2=t and t1t2, see Figure 1. We call Qv(s,t1,t2) single tail Q graph if t2=0, simplifies as Qv(s,t), and double tails Q graph if t21. Clearly, Q graph is formed by identifying two vertices of cycle and path.

    Figure 1.  Double tail Qv(s,t1,t2) graph.

    In this paper, we determine the burning number of the Q graph and characterize Q graph by the burning number. For a simple undirected graph G which has vertex set V(G) and edge set E(G) and vV(G), the eccentricity of v is ecc(v)=max{d(v,u):uV(G)}, where d(v,u) represents the distance of the shortest path from v to u. The radius and diameter of G are defined as r(G)=min{ecc(v):vV(G)} and d(G)=max{ecc(v):vV(G)}, respectively. Here we list some known results.

    A spider graph Sv is a tree with exactly one vertex v of degree strictly greater than two and v is the center of the spider. A spider Sv with r arms can be thought of as Pa1Pa2Par, where the Pai are edge-disjoint paths that share one common end vertex v. Then, Sv is also denoted as Sv(a1,a2,,ar).

    Proposition 1.3. [7] Let G=Sv(a1,a2,,ar) be a spider graph of order n. Then, b(G)n.

    Proposition 1.4. [3] For a graph G, we have that b(G)=min{b(T):TisaspanningtreeofG}.

    Proposition 1.5. [3] Burning a graph G in k steps is equivalent to finding a rooted tree partition into k trees T1,T2,,Tk, with heights at most (k1),(k2),,0, respectively, such that for every 1i,jk, the distance between the roots of Ti and Tj is at least |ij|.

    Proposition 1.6. [3] For any graph G with radius r and diameter d, we have that (d+1)1/2b(G)r+1.

    Proposition 1.7. [3] Let H be an isometric subgraph of a graph G such that, for any node xV(G)V(H), and any positive integer r, there exists a node fr(x)V(H) for which Nr[x]V(H)NHr[fr(x)]. Then, b(H)b(G).

    Proposition 1.8. [12] Let G=Pa1Pa2 with a1a21 and J(t)={(t22,2)} for integer t2. Then,

    b(G)={a1+a2+1,If(a1,a2)J(t);a1+a2,Otherwise.

    Proposition 1.9. [12] Let G=Pa1Pa2Pa3 with a1a2a31. Then,

    b(G)={a1+a2+a3+1,If(a1,a2,a3)J1J2J3J4J5;a1+a2+a3,Otherwise.

    Let Ji for 1i5 satisfy the following conditions.

    D1={(2,2)},D2={(3,2)},D3={(1,1),(3,3),(4,2),(5,5)},D4={(2,1),(4,1),(4,3),(4,4),(6,1),(6,4),(6,5),(6,6),(7,7),(8,4),(8,6),(10,4)},D5={(11,10,4),(13,11,1),(11,11,3),(22,13,1),(19,13,4),(17,13,6),(15,13,8),(13,13,10),(17,15,4),(15,15,6),(30,15,4),(28,15,6),(26,15,8),(19,15,15),(28,17,4),(26,17,6),(17,17,15),(26,19,4),(43,17,4),(41,17,6),(30,17,17),(41,19,4),(30,30,4),(58,19,4)},J1={(a1,a2,a3):(a2,a3)D1,a1+a2+a3=t23 for integer t},J2={(a1,a2,a3):(a2,a3)D1D2,a1+a2+a3=t22 for integer t},J3={(a1,a2,a3):(a2,a3)3i=1Di,a1+a2+a3=t21 for integer t},J4={(a1,a2,a3):a3=2 or (a2,a3)4i=1Di,a1+a2+a3=t2 for integer t},J5=D5{11,11,2}.

    In this section, we determine the burning number of the single tail Q graphs Qv(s,t) and confirm the Bonato's conjecture. Further more, we characterize the single tail Q graph by the burning number. Now we give some useful lemmas.

    Lemma 2.1. Let G=Qv(s,t) be a single tail Q graph with order n, we have that n1b(G)n.

    Proof. Since G has a hamiltonian path, by Theorem 1.1, the upper bound is clear. Let v be the 3-degree vertex and (x1,x2,,xk) be an optimal burning sequence of G. Let Ti denote the derived subtree with the root xi and a height that does not exceed ki in G. If no Ti is spider, then |Nki[xi]|2(ki)+1 for 1ik. Thus, we have ki=1(2(ki)+1)=k2n, so b(G)=kn. If there exist a root subtree, named T1, is the spider Sv(a1,a2,a3) for d(v,x1)=l. Then, Nk1[x1]3(k1)+1l. Combine this with the fact |Nki[xi]|2(ki)+1 for 2ik, we have

    3(k1)+1l+ki=2(2(ki)+1)=(3(k1)+1l)+(2k3)++1=k2+k1l.

    Considering k2+k1ln and l0, gives us k(n+5/4)1/21/2n1. This completes the proof.

    Lemma 2.2. Let G=Qv(s,t) be a single tail Q graph with b(G)=k. Then, |V(G)|k2+k1.

    Proof. Suppose v is 3-degree vertex of G and (x1,x2,,xk) is an optimal burning sequence of G. In order to let G contain as many vertices as possible, the following conditions must holds:

    (1) x1=v;

    (2) ki=1Nki[xi]=V(G);

    (3) Nki[xi]Nkj[xj]= for 1ijk.

    Consider d(xi)=2 for 2ik, so |Nki[xi]|=2k2i+1. Combine |Nk1[v]|=3k2, the maximum number of |V(G)| is Σki=1|Nki[xi]|

    =(3k2)+(2k3)++(2k2i+1)++3+1=k2+k1.

    By Lemma 2.1 and 2.2, we get the following corollaries.

    Corollary 2.3. Let G=Qv(s,t) be a single tail Q graph with order k2+r for 1r2k+1. Then, kb(G)k+1.

    Corollary 2.4. Let G=Qv(s,t) be a single tail Q graph with order k2+r for kr2k+1. Then, b(G)=k+1.

    The following we determine the exact value of the burning number of single tail Q graph b(Qv(s,t)) with order k2+r for 1rk1.

    Lemma 2.5. Let G=Qv(s,t) be a single tail Q graph with order k2+r for 1rk1. If sk2 or tk2, then b(G)=k+1.

    Proof. When sk2, consider Cs+1 is an isometric subgraph of G with |V(Cs+1)|k2+1. Then, by Theorem 1.1, b(Cs+1)=|V(Cs+1)|k+1. Further more, we find that for any vertex xV(G)V(Cs+1) and any positive integer r, Nr[x]V(Cs+1)NCs+1r[v]. By Proposition 1.7 and Corollary 2.3, we get k+1b(Cs+1)b(G)k+1. So, b(G)=k+1. Similarly, since Pt+1 is an isometric subgraph of G with tk2, we can show that b(G)=k+1.

    Lemma 2.6. Let G=Qv(s,t) be a single tail Q graph with order k2+k1. Then,

    b(G)={k,If2k2sk21butsk23,2k;k+1,Otherwise.

    Proof. First, consider the case for 2s2k3. Since d(G)=s2+t=s2+k2+k1s1=k2+k1s+12k2+k1(k1)=k2, by Proposition 1.6 and Corollary 2.3, we have b(G)=k+1. For the case s=k23 or 2k, we get b(G)=k+1. In fact, if not, assume b(G)=k. Then, by Lemma 2.2, we have x1=v. Let H=GNk1[v], then b(H)=k1. However, notice H=PsPt with s=s(2k2)=k22k1 or 2, t=t(k1)=2 or k22k1, by Proposition 1.8, we have b(H)=k, which contradicts to b(H)=k1. So, we get b(G)=k+1.

    Second, consider the case for 2k2sk21 and let x1=v, and H=GNk1[v]. Then, we have H=PsPt with s=s(2k2) and t=t(k1). Further more, notice that s+t=k22k+1 and sk22k1 or 2, so we have t2 or k22k1. Then, by Proposition 1.8, we get b(H)=k1 and b(G)1+b(H)=k. Combine this with Corollary 2.3, and we get b(G)=k.

    Lemma 2.7. Let G be single tail Q graph with order k2+r for 1rk2. If 2s2k3, then

    b(G)={k,Ifs/2r;k+1,Ifs/2<r.

    Proof. First consider the case for s/2<r. Since d(G)=s2+t=k2+rs1+s2=k2+rs/21>k21, d(G)k2. By Proposition 1.6 and Corollary 2.3, we have b(G)d(G)+1=k+1 and thus b(G)=k+1. Now consider the case for s/2r.

    If s is odd, let (s1)/2=s and x1=uV(Pt+1) such that d(v,u)=ks2. Then, V(Cs+1)Nk1[u] and thus GNk1[u]=Pt with t=k2+r(2s+1)(d(u,v)+1)(k1)=k2+rs2k+1k2+rr2k+1=(k1)2. So, we have b(Pt)=tk1. Thus, b(G)1+b(Pt)k. Combine this with Corollary 2.3 and we get b(G)=k. If s is even, let s/2=s and x1=uV(Pt+1) such that d(v,u)=ks1. Then, V(Cs+1)Nk1[u]. Similarly, since GNk1[u]=Pt with t=k2+r2s(d(u,v)+1)(k1)=k2+rs2k+1k2+rr2k+1=(k1)2, we have b(Pt)tk1 and thus b(G)k. So b(G)=k.

    Lemma 2.8. Let G be single tail Q graph with order k2+r for 1rk2. If 2k2sk21, then b(G)=k.

    Proof. Let x1=v and H=GNk1[v]. If tk, denote t=t(k1), s=s(2k2). Then, H=PsPt with s+t=k2+r13(k1)k22k<(k1)2. By Proposition 1.8, we have that b(H)k1. Thus, b(G)k. If tk1, V(Pt+1)Nk1(v) and H=Ps with s=s(2k2). Note that sk21, so s(k1)2. Thus, b(H)k1. So, b(G)k.

    To summarize the above discussion, we get following results.

    Theorem 2.9. Let G=Qv(s,t) be a single tail Q graph with order k2+r with 1r2k+1. Then,

    b(G)={k,If1rk2,tk21,2s2k3ands/2r;or1rk2,tk21and2k2sk21;orr=k1,tk21,2k2sk21andsk23,2k;k+1,Otherwise.

    In this section, we determine the burning number of the double tails Q graph G=Qv(s,t1,t2) and confirm the Bonato's conjecture. Further more, we characterize the double tails Q graph by the burning number. First, we give some lemmas.

    Lemma 3.1. Let G=Qv(s,t1,t2) be a double tails Q graph with order k2+r with 1r2k+1. Then, kb(G)k+1.

    Proof. Notice that Sv(s,t1,t2) is a spanning tree of G. Then, by Propositions 1.3 and 1.4, we get b(G)b(Sv(s,t1,t2))k2+r=k+1. Now we show b(G)k. Let (x1,x2,,xp) be an optimal burning sequence of G, Ti(1ip) be the derived subtrees with the root xi, and the height of Ti be less than or equal to pi. If there is no spider in Ti(1ip), then |Npi[xi]|2(pi)+1 for 1ip. Thus, pi=1(2(pi)+1)=p2k2+r and b(G)=pk2+r=k+1>k. If there exist a spider in Ti(1ip), without loss generality, assume T1 is a spider. Suppose T1 is three-arms spider Sv(a1,a2,a3) such that d(v,x1)=l. Then, Np1[x1]3(p1)+1l. Similar to Lemma 2.1, we have b(G)=pk2+r1k. Suppose T1 is a four-arms spider Sv(a1,a2,a3,a4) such that d(v,x1)=m. Then, we have |Np1[x1]|4(p1)+12m. Considering |Npi[xi]|2(pi)+1 for 2ip, we get

    |Np1[x1]|+pi=2(2(pi)+1)(4(p1)+12m)+(2(p2)+1)++3+1(4(p1)+1)+(2p3)++1=p2+2p2.

    Considering pi=1Npi[xi]=V(G), we have p2+2p2k2+r. Thus, pk2+r+31k.

    Clearly, by Lemma 3.1, the burning number of double tails Q graph Qv(s,t1,t2) with order k2+r for 1r2k+1 is either k+1 or k. Further more, by the proof, we know that the order of Qv(s,t1,t2) with the burning number k is not more than k22k+2.

    Claim 3.2. Let Sv(a1,a2,,aΔ) be a spider and G be a path-forest. If H=Sv(a1,a2,,aΔ)G and b(H)=p, then |V(H)|(p1)2+Δ(p1)+1.

    Lemma 3.3. Let G=Qv(s,t1,t2) be a double tails Q graph with order p2+r for r2p1. Then, b(G)=p+1.

    Proof. Notice that every spanning tree of G is a union of spider and path-forest and |V(G)|=p2+rp2+2p1=(p1)2+4(p1)+2. Thus, by Claim 3.2, have b(G)=p+1.

    Lemma 3.4. Let G=Qv(s,t1,t2) be a double tails Q graph with order p2+r for 1r2p2. If sp2 or t=t1+t2p2 or t1+s2p2, then b(G)=p+1.

    Proof. Clearly, Cs+1 is an isometric subgraph of G. If sp2, then |V(Cs+1)|p2+1. By Theorem 1.1 we have b(Cs+1)=|V(Cs+1)|p+1. Further more, for any vertex xV(G)Cs+1 and positive integer r, we have Nr[x]V(Cs+1)NCs+1r[v]. By Proposition 1.7 and Corollary 2.1, we get p+1b(Cs+1)b(G)p+1. So, b(G)=p+1. Similarly, notice that Pt+1,Qv(s,t1) are also isometric subgraphs of G. Consider tp2 and t1+s2p2. Then, by Proposition 1.7 and Corollary 2.1, we get p+1b(Pt+1)b(G)p+1 and p+1b(Qv(s,t1))b(G)p+1. Thus, b(G)=p+1.

    Now we go on proposing claims.

    Claim 3.5. Let G=Qv(s,t1,t2) be a double tails Q graph with order p2+r for 2p2m1r2p2m and mp1. If b(G)=p, then there exists one i{1,2,,m} such that xiNmi[v].

    Proof. Suppose (x1,x2,,xp) is an optimal burning sequence of G with order p2+r for 2p2m1r2p2m and mp1. Assume that all xiNmi[v] for i=1,2,,m. Let H be the subgraph of G which induced by Np1[x1]Np2[x2]Npm[xm]. Now we show b(GH)>pm, which implies that b(G)>p. This contradicts to b(G)=p.

    If vH, suppose vNpj[xj] for j{1,2,,m}, combine xjNmj[v] we have |Npj[xj]|4(pj)2(mj+1)+1=4p2m2j1. Consider |Npi[xi]|2(pi)+1 for ij, we get

    |H|mi=1(2p2i+1)+|Npj[xj]|(2p2j+1)=2p+2mpm22m2.

    Thus, |V(GH)|(pm)2+1. Considering that GH is a path forest, then b(GH)>pm, a contradiction.

    If vH, |Npi[xi]|2(pi)+1 for i{1,2,,m}. Thus, |V(GH)|p2+2p(2m+1)mi=1(2p2i+1)=(pm)2+2(pm)1=(pm1)2+4(pm1)+2. By Claim 3.2, we have that b(GH)>pm, a contradiction.

    Claim 3.6. Let G=Qv(s,t1,t2) be a double tails Q graph with order p2+2p2. If b(G)=p, then 2p2sp21, p1t2t1, t1+t2p21, and t1+s2p21.

    Proof. Suppose (x1,x2,,xp) is an optimal burning sequence of G. Then, by Lemma 3.4, the upper bounds of s,t1+t2 and t1+s2 are clear. Here we only show the lower bound. Consider r=2p2 and b(G)=p, and then by Claim 3.5, we know that x1=v. Now assume that s2p3 or t2t1p2. Then, |Np1[v]|4p4 and thus |V(GNp1[v])|p22p+2. Consider that GNp1[v] is path forests. By Proposition 1.9, we get b(GNp1[v])p and thus b(G)p+1, a contradiction.

    Lemma 3.7. Let G=Qv(s,t1,t2) be a double tails Q graph with order p2+r for 2p3r2p2. If 2p2sp21, p1t2t1, t1+t2p21, and t1+s2p21, then

    {b(G)=p+1,Ifst20and(s,t1,t2)J1J2J3J4J5;orst2=0,r=2p2andmax{s,t1}=p22p1.b(G)=p,Otherwise.

    where s=s(2p2),t1=t1(p1),t2=t2(p1).

    Proof. Suppose s=s(2p2),t1=t1(p1),t2=t2(p1) and we distinguish cases to complete the proof.

    Case 1. st20

    If (s,t1,t2)J1J2J3J4J5, we let x1=v and H=GNp1[x1]. Notice that |Np1[x1]|=4p3 and H=PsPt1Pt2. Then, p22p|H|p22p+1. By Proposition 1.9, we have b(H)=p1. Thus, b(G)b(H)+1=p. Combining this with Lemma 3.1, we get b(G)=p. If (s,t1,t2)J1J2J3J4J5. Assume b(G)=p. Then, by Claim 3.5, we have x1=v. Let H=GNp1[x1]. Then, b(H)=p1. Notice that H=PsPt1Pt2. Therefore, by Proposition 1.9, we have b(H)=p, a contradiction. Thus, b(G)=p+1.

    Case 2. st2=0

    If max{s,t1}p22p1 or r2p2, let x1=v and H=GNp1[x1]. Consider |Np1[x1]|=4p3 and H is 2-path forests. Thus, by Proposition 1.8, we have b(H)=p1. Then, b(G)b(H)+1=p. Combining this with Lemma 3.1, we have b(G)=p. If max{s,t1}=p22p1 and r=2p2, assume b(G)=p rather than p+1, by Claim 3.5 we have x1=v and thus b(H)=p1 for H=GNp1[x1]. However, notice r=2p2, we have |V(H)|=p22p+1 and H=P2Pp22p1. By Proposition 1.8, we have b(H)=p, a contradiction. Thus, b(G)=p+1.

    Lemma 3.8. Let G=Qv(s,t1,t2) be a double tails Q graph with order p2+2p4. If 2p2sp21, p1t2t1, t1+t2p21 and t1+s2p21. Then,

    {b(G)=p+1,Ifs=2pandt2{p+1,p+2};ort2=p+1,t1=p+2;b(G)=p,Otherwise;

    Proof. Clearly, t1+t2+s+1=p2+2p4. Now we distinguish cases to discuss.

    Case 1. s2p.

    Subcase 1.1. t2p+1.

    Let x1=v and H=GNp1[v]. Then, |V(H)|=(p1)22 and H=PsPt1Pt2 with s=s(2p2), t1=t1(p1) and t2=t2(p1)2. Let y=min{s,t1}, and then it is clear that (y,t2)D1D2. Then, by Proposition 1.9 and Lemma 3.1, we have b(H)=p1 and thus b(G)=p.

    Subcase 1.2. t2=p+1 but t1p+2.

    If t1=p+1, let x1V(Pt2) with d(x1,v)=1 and H=GNp1[x1], |V(H)|=(p1)2 and H=PsP3P1 with s=s(2p4). Notice |V(H)|=(p1)2>4, we have p4 and thus s5. Clearly, (s,3,1)J1J2J3J4J5. By Lemma 3.1, we have b(H)=p1 and thus b(G)=p.

    Now consider the case for t1p+3. If s=2p+1, then t1=p2p7. Let x1V(Pt2) with d(x1,v)=1. Let H=GNp1[x1]. Then, |V(H)|=(p1)2 and H=PsPt1Pt2 with s=s2(p2)=5, t1=t1(p2)5 and t2=t2p=1. Notice (1,5)4i=1Di. Hence (s,t1,t2)J1J2J3J4J5, and by Proposition 1.9 and Lemma 3.1, we have b(H)=p1. Thus b(G)=p. If s2p+1, let x1=v and H=GNp1[v], and consider |Np1[x1]|=4p3. Then, |V(H)|=(p1)22. Let H=PsPt1Pt2 with s=s(2p2)2,3, t1=t1(p1)4, t2=t2(p1)=2 and y=min{s,t1}. Notice (y,t2)D1D2, so by Proposition 1.9 and Lemma 3.1, we have b(H)=p1 and thus b(G)=p.

    Subcase 1.3. t1=p+2 and t2=p+1.

    In this case we give a proof to show b(G)=p+1 by contradiction. Assuming that b(G)=p, by Claim 3.5, we have that x1N1[v] or x2=v.

    If x1N1[v], let H=GNp1[x1]. Then, b(H)=p1. In fact, we can prove b(H)=p to derive contractions. If x1=v, then H=PsP2P3 with |V(H)|=(p1)22 and s=s2(p1). Notice |V(H)|5, we have p4 and s=p2+2p4(p+2)(p+1)1=p28. Thus, s=p282(p1)2, by Proposition 1.9, we have b(H)=p.

    If x1V(Cs+1) with d(x1,v)=1, then H=PsP4P3 with |V(H)|=(p1)2 and s=s2(p1). Considering |V(H)|7, we have p4, then s2. Let y=min{s,4} and consider {y,3}D1D2D3D4. Then, by Proposition 1.9, we have b(H)=p. If x1V(Pt1) with d(x1,v)=1, then H=PsP2P3 with |V(H)|=(p1)2 and s=s2(p2)2. By Proposition 1.9, we have b(H)=p. Similarly, if x1V(Pt2) with d(x1,v)=1, then we also can prove b(H)=p.

    If x2=v, let H=GNp2[x2]Np1[x1], and considering |GNp2[x2]|=p22p+3, we get b(H)=p2. Now we show b(H)=p1 to derive contractions. Consider |Np1[x1]|=2(p1)+1 and |GNp2[x2]|=p22p+3t1(p2)+t2(p2)=7, such that we have p4. Then, |Np1[x1]|7. So, x1V(Cs+1) and H=PsP4P3 with |V(H)|=(p2)2 and s2 for (p2)2>7. By Lemma 3.1 and Proposition 1.9, we have b(H)=p1.

    Case 2. s=2p.

    Subcase 2.1. t2{p+1,p+2}.

    Similarly, we have a contradiction to show b(G)=p+1. Assume b(G)=p, by Claim 3.5, we have that x1N1[v] or x2=v. Let H=GNp1[x1] or GNp2[x2]Np1[x1]. Then, b(H)=p1 or p2. Now we show b(H)=p or p1 to derive contractions, respectively.

    If x1=v, then |V(H)|=(p1)22 and H=P2P2Pt1 or P2P3Pt1 with t1=t1(p1)2. By Proposition 1.9, b(H)=p. If x1V(Cs+1) with d(x1,v)=1, then |Np1[x1]|=4p5, |V(H)|=(p1)2 and H=P2P3Pt1 or P2P4Pt1 with t1=t1(p2)3. By Proposition 1.9, we also get b(H)=p, contradiction. Similarly, we can show b(H)=p for other cases x1V(Pt1) with d(x1,v)=1 and x1V(Pt2) with d(x1,v)=1, details are omitted.

    If x2=v, let H=GNp2[x2]Np1[x1], and consider b(G)=p. Then, b(H)=p2. Clearly, |V(H)|(p2)2. Further more, we get |V(H)|=(p2)2. If |V(H)|>(p2)2, then b(H)p1, which contradicts to b(H)=p2. Since t2(p2)4, s2(p2)=4, we have x1V(Pt1). Consider P4P3 or P4P4H, such that p5. By Proposition 1.9, we have b(H)=p1, a contradiction.

    Subcase 2.2. t2{p+1,p+2}.

    In this case, we let x1=v and H=GNp1[x1]. Clearly, V(H)=(p1)22 and H=Pt1Pt2P2, satisfying t1=t1(p1), t2=t2(p1){2,3}. By Proposition 1.9 and Lemma 3.1, we have b(H)=p1 and thus b(G)=p.

    Lemma 3.9. Let G=Qv(s,t1,t2) be a double tails Q graph with order p2+2p5. If 2p2sp21, p1t2t1, t1+t2p21 and t1+s2p21, then b(G)=p.

    Proof. If t2=p+1, we let x1V(Pt2) such that d(v,x1)=1 and H=GNp1[x1]. Consider |V(H)|=(p1)21 and H=PsPt1P1 with s=s(2p4)2 and t1=t1(p2)3. Let y=min{s,t1}. Since (y,1)D1D2D3, by Proposition 1.9 and Lemma 3.1, we get b(H)=p1 and thus b(G)=p.

    If t2p+2 or p1t2p, let x1=v and H=GNp1[v]. Then, |V(H)|=(p1)23 and H=PsPt1Pt2 with s=s(2p2), t1=t1(p1) and t2=t2(p1)2. When st2=0, by Proposition 1.8, we get b(H)=p1. If st20, let y=min{s,t1} and notice (y,t2)D1. Then, by Proposition 1.9 and Lemma 3.1, we have b(H)=p1 and thus b(G)=p.

    Lemma 3.10. Let G=Qv(s,t1,t2) be a double tails Q graph with order p2+r. If r2p6, 2p2sp21, p1t2t1, t1+t2p21 and t1+s2p21, then b(G)=p.

    Proof. Let x1=v and H=GNp1[v] such that b(H)p1. Clearly, since |Np1[v]|=4p3 and H=PsPt1Pt2 for s=s(2p2), t1=t1(p1) and t2=t2(p1), |V(H)|(p1)24. By Propositions 1.8 and 1.9, we get b(H)p1. This completes the proof.

    By Lemmas 3.7–3.10, we obtain the burning number of Qv(s,t1,t2) with order p2+r for 1r2p3 when 2p2sp21, p1t2t1, t1+t2p21 and t1+s2p21. Now, we get following results.

    Theorem 3.11. Let G=Qv(s,t1,t2) be a double tails Q graph with order p2+r for 1r2p+1. If r2p1 or 1r2p2 and sp2 or 1r2p2 and t1+t2p2 or 1r2p2 and t1+s2p2, then b(G)=p+1.

    Following Theorem 3.11, we consider the case for 2p2sp21, t1+t2p21 and t1+s2p21.

    Theorem 3.12. Let G=Qv(s,t1,t2) be a double tails Q graph with order p2+r for 1r2p2 and 2p2sp21, t1+t2p21 and t1+s2p21. If s=s(2p2),t1=t1(p1),t2=t2(p1), then b(G)=p.

    At the end of this section, we discuss cases for s2p3, t2p2 and t1p2, respectively.

    Theorem 3.13. Let G=Qv(s,t1,t2) be a double tails Q graph with order p2+r. If 1r2p3, 2p2sp21, t2p2, t1p1, t1+t2p21 and t1+s2p21, then

    {b(G)=p,Ifrp2;orrp1,t2=rp+1ands{p23,2p};orrp1andt2rp+2;b(G)=p+1,Otherwise;

    Proof. Clearly, G=Qv(s,t1) is an isometric subgraph of G and for any uV(GG) and any position integer r, NGr[u] V(G)NGr[v]. Thus, by Proposition 1.7, we get b(G)b(G).

    First, consider the case for rp1 and 1t2rp. Then, |V(G)|=nt2p2+p. By Corollary 2.4, we know that p+1=b(G)b(G). Thus we get b(G)=p+1. If rp1, t2=rp+1, then |V(G)|=p2+p1. When s{p23,2p}, by Lemma 2.6, we have b(G)=p. Further, by Lemma 2.2 we know that x1=v. Notice t2p2, so V(Pt2)Np1[v] and thus we get b(G)=p. When s{p23,2p}, by Lemma 2.6, we have b(G)=p+1. Thus b(G)=p+1. Next, we consider two cases for rp1, t2rp+2 and rp2. Notice that |V(G)|p2+p2 in these two cases. By Lemma 2.8, we know that x1=v and b(G)=p. By combining this with V(Pt2)Np1[v] for t2p2, we have b(G)=p.

    Theorem 3.14. Let G=Qv(s,t1,t2) be a double tails Q graph with order p2+r. If 1r2p3, 2p2sp21, t1p2 and t1+s2p21, then b(G)=p.

    Proof. Let x1=v. Then, it is clear that V(Pt+1)Np1[v]. Suppose Ps=GNp1[x1] with s=s(2p2)p22p+1. Then, by Theorem 1.1 we have b(Ps)p1. Thus b(G)b(Ps)+1p. By combining this with Lemma 3.1, we have b(G)=p.

    Theorem 3.15. Let G=Qv(s,t1,t2) be a double tails Q graph with order p2+r. If 1r2p3, s2p3, t=t1+t2p21 and t1+s2p21. Then,

    {b(G)=p+1,Ift=p21,s=2p3andt2=p+1;b(G)=p,Otherwise.

    Proof. Now we distinguish three cases to complete the proof.

    Case 1. t2s2.

    Let x1V(Pt1+1) such that d(x1,v)=p1s2. Clearly, V(Cs+1)Np1[x1] and V(Pt2)Np1[x1] for t2s2. Thus GNp1[x1]=Pt1 with t1=tt2d(x1,v)(p1)1p22p, by Theorem 1.1 and Lemma 3.1 we have b(Pt1)p1. Thus pb(G)b(Pt1)+1p and we get b(G)=p.

    Case 2. t2>s2 but t2s2+2 or tp22.

    Let x1V(Pt1+1) such that d(x1,v)=p1s2, it is clear that V(Cs+1)Np1[x1] and GNp1[x1]=Pt1Pt2 with t1=t1d(x1,v)(p1), t2=t2+d(x1,v)(p1). If t2s2+2, then t22 and t1+t2p22p+1, by Proposition 1.8, we have b(Pt1Pt2)=p1. Thus b(G)=p. If tp22, then t1+t2p22p. By Proposition 1.8, we know that b(Pt1Pt2)=p1. Combined with Lemma 3.1, we get b(G)=p.

    Case 3. t2=s2+2 and t=p21.

    If s2p4, let x1V(Pt1) such that d(x1,v)=p2s2. Clearly V(Cs+1)Np1[x1] and Qv(s,t1,t2)Np1[x1]=Pt1Pt2 with t1=t1d(x1,v)(p1)=p22p and t2=t2+d(x1,v)(p1)=1. By Proposition 1.8 and Lemma 3.1, we have b(Pt1Pt2)=p1 and thus b(G)=p.

    If s=2p3, then t2=p+1 and we can show b(G)=p+1. If not, assume b(G)=p and (x1,x2,,xp) is an optimal burning sequence for G. Notice H=Pt1+t2+1 is an isometric subgraph of G and for any vertex uV(GH) and any position integer r, NGr[u]V(H)NHr[v]. By Proposition 1.7, we get b(H)b(G)=p. Consider |V(H)|=p2, this means {x1,x2,,xp}V(H) and V(Cs+1)Npi[xi] for some i. Further, by r(Cs+1)=s2=p1 we know V(Cs+1)Np1[x1] for x1=v, we find GNp1[v]=P2Pp22p1. By Proposition 1.8, we get b(P2Pp22p1)=p and thus b(G)=p+1, a contradiction. Then, b(G)=p+1.

    The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.

    The authors would like to thank anonymous reviewers for their valuable comments and suggests to improve the quality of the article. This work was supported by NSFC No.12371352 and Qinghai provincial basic research project No.2022-ZJ-753.

    The authors declare no conflicts of interest.



    [1] J. A. Bondy, U. S. R. Murty, Graph theory with applications, J. Oper. Res. Soc., 28 (1977), 237–238. http://dx.doi.org/10.1057/jors.1977.45 doi: 10.1057/jors.1977.45
    [2] A. Bonato, J. Janssen, E. Roshanbin, Burning a graph as a model of social contagion, In: Algorithms and models for the web graph, 2014. https://doi.org/10.1007/978-3-319-13123-8_2
    [3] A. Bonato, J. Janssen, E. Roshanbin, How to burn a graph, Internet Math., 12 (2016), 85–100. http://dx.doi.org/10.1080/15427951.2015.1103339 doi: 10.1080/15427951.2015.1103339
    [4] A. Bonato, J. Janssen, E. Roshanbin, Burning a graph is hard, Discrete Appl. Math., 232 (2017), 73–87. http://dx.doi.org/10.1016/j.dam.2017.07.016 doi: 10.1016/j.dam.2017.07.016
    [5] N. Alon, P. Pralat, N. Wormald, Cleaning regular graphs with brushes, Siam. J. Discrete Math., 23 (2008), 233–250. http://dx.doi.org/10.1137/070703053 doi: 10.1137/070703053
    [6] A. Barghi, P. Winkler, Firefighting on a random geometric graph, Random. Struct. Algor., 46 (2015), 466–477. http://dx.doi.org/10.1002/rsa.20511 doi: 10.1002/rsa.20511
    [7] A. Bonato, T. Lidbetter, Bounds on the burning numbers of spiders and path-forests, Theor. Comput. Sci., 794 (2019), 12–19. http://dx.doi.org/10.1016/j.tcs.2018.05.035 doi: 10.1016/j.tcs.2018.05.035
    [8] K. A. Sim, T. S. Tan, K. B. Wong, On the burning number of generalized Petersen graphs, Bull. Malays. Math. Sci. Soc., 41 (2018), 1657–1670. http://dx.doi.org/10.1007/s40840-017-0585-6 doi: 10.1007/s40840-017-0585-6
    [9] D. Mitsche, P. Pralat, E. Roshanbin, Burning number of graph products, Theor. Comput. Sci., 746 (2018), 124–138. http://dx.doi.org/10.1016/j.tcs.2018.06.036 doi: 10.1016/j.tcs.2018.06.036
    [10] D. Mitsche, P. Pralat, E. Roshanbin, Burning graphs: A probabilistic perspective, Graph Combinator., 33 (2017), 449–471. http://dx.doi.org/10.1007/s00373-017-1768-5 doi: 10.1007/s00373-017-1768-5
    [11] H. Q. Liu, R. T. Zhang, X. L. Hu, Burning number of theta graphs, Appl. Math. Comput., 361 (2019), 246–257. http://dx.doi.org/10.1016/j.amc.2019.05.031 doi: 10.1016/j.amc.2019.05.031
    [12] H. Q. Liu, X. J. Hu, X. L. Hu, Burning numbers of path forests and spiders, Bull. Malays. Math. Sci. Soc., 44 (2021), 661–681. http://dx.doi.org/10.1007/s40840-020-00969-w doi: 10.1007/s40840-020-00969-w
  • This article has been cited by:

    1. Kai An Sim, Kok Bin Wong, On the cooling number of the generalized Petersen graphs, 2024, 9, 2473-6988, 36351, 10.3934/math.20241724
  • Reader Comments
  • © 2024 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(1427) PDF downloads(88) Cited by(1)

Figures and Tables

Figures(1)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog