Elsevier

Applied Mathematics and Computation

Volume 336, 1 November 2018, Pages 94-106
Applied Mathematics and Computation

On the edge-Szeged index of unicyclic graphs with given diameter

https://doi.org/10.1016/j.amc.2018.04.077Get rights and content

Abstract

Given a connected graph G, the edge-Szeged index Sze(G) is defined as Sze(G)=e=uvEmu(e)mv(e), where mu(e) and mv(e) are, respectively, the number of edges of G lying closer to vertex u than to vertex v and the number of edges of G lying closer to vertex v than to vertex u. In this paper, some extremal problems on the edge-Szeged index of unicyclic graphs are considered. All the n-vertex unicyclic graphs with a given diameter having the minimum edge-Szeged index are identified. Using a unified approach we identify the n-vertex unicyclic graphs with the minimum, second minimum, third minimum and fourth minimum edge-Szeged indices.

Introduction

In this paper, we consider connected simple and finite graphs, and refer to Bondy and Murty [2] for notations and terminologies used but not defined here.

Let G=(VG,EG) be a graph with vertex set VG={v1,v2,,vn} and edge set EG. We call n ≔ |VG| the order of G and |EG| the size of G. Denote by Pn, Cn and Kn a path, a cycle and a complete graph of order n, respectively. The set of neighbors of a vertex v in G is denoted by NG(v) or simply N(v). The degree dG(v) of a vertex v is equal to the number of neighbors of v. An edge uv is called a pendant edge if one of its endpoints is of degree 1. By Guv we denote the graph obtained from G by deleting an edge uv ∈ EG. (This notation is naturally extended if more than one edge are deleted.) Similarly, G+uv is obtained from G by adding an edge uvEG. The distance, dG(u, v) (or d(u, v) for short), between vertices u and v of G is the length of a shortest u, v-path in G.

Recently, in organic chemistry, topological indices have been found to be useful in chemical documentation, structure–property relationships, structure–activity relationships and pharmaceutical drug design. These indices include Wiener index [43], Randić index [28], [29], Hosoya index [17], [18], graph energy [36], matching energy [15], [16], [27], the HOMO-LUMO index [32] and Zagreb index [8] and so on.

Among all the topological indices, the most well-known is the Wiener index and it was studied extensively. The Wiener index (or transmission) W(G) of G is the sum of distances between all pairs of vertices of G, that is, W(G)={u,v}VGdG(u,v).This distance-based graph invariant was in chemistry introduced back in 1947 [43] and in mathematics about 30 years later [12]. Nowadays, the Wiener index is an extensive studied graph invariant; see the surveys [7], [10]. A collection of recent papers dedicated to the investigations of the Wiener index [23], [25], [31].

Given an edge e=uv of a graph G, the sets N0(e), Nu(e) and Nv(e) are defined to be the set of vertices equidistant from u and v, the set of vertices whose distance to vertex u is smaller than the distance to vertex v and the set of vertices closer to v than u, respectively. Put nu(e)=|Nu(e)|,nv(e)=|Nv(e)| and n0(e)=|N0(e)|. Obviously, Nu(e)∪Nv(e)∪N0(e) is a partition of VG with respect to e, where N0(e)= if G is bipartite. Thus, one has nu(e)+nv(e)+n0(e)=|VG|.

Gutman [14] introduced the Szeged index Sz(G) of a graph G as Sz(G)=e=uvEGnu(e)nv(e).Furthermore, in order to involve also those vertices that are at equal distance from the endpoints of an edge, Randić [37] proposed the revised Szeged index Sz*(G) of a graph G as follows: Sz*(G)=e=uvEG(nu(e)+n0(e)2)(nv(e)+n0(e)2).For more detailed information on Szeged index and revised Szeged index, one may be referred to the recent work [40], [41] and the references therein.

Since Sz(T)=W(T) holds for any tree T, a lot of research has been done on the relation between the (revised) Szeged index and the Wiener index on general graphs. On the one hand, the difference between Sz(G) and W(G) was investigated systematically in [1], [4], [5], [9], [20], [21], [22], [34], [35], [47]; On the other hand, the quotient between (revised) Szeged index and the Wiener index was also studied in [26], [46].

Given an edge e=uvEG, the distance between the edge e and the vertex w, denoted by d(e, w), is defined as d(w,e)=min{d(u,w),d(v,w)}.Then, let Mu(e|G)={xEG:d(u,x)<d(v,x)},Mv(e|G)={xEG:d(v,x)<d(u,x)},M0(e|G)={xEG:d(u,x)=d(v,x)}.For simplicity, let Mu(e)=Mu(e|G),Mv(e)=Mv(e|G) and M0(e)=M0(e|G). Put mu(e)=|Mu(e)|,mv(e)=|Mv(e)| and m0(e)=|M0(e)|. Then, the edge-Szeged index [15] of G is defined as Sze(G)=e=uvEGmu(e)mv(e).

In the case of trees, in view of Gutman and Ashrafi [15], we know that Sze(T)=W(T)(n1)2. Note that the Wiener index of trees with some given parameters have been studied extensively. Hence, it is interesting for us to study the edge-Szeged index of connected graphs with cycles.

The edge-Szeged index is also a distance-based graph invariant, it attracts more and more researchers’ attention. On the one hand, one focuses on finding method to compute the edge-Szeged index of some type of graphs; see [13], [33], [42], [45]; On the other hand, the researchers are also interested in determining the graph with the maximum or minimum edge-Szeged index among some type of graphs; see [3], [6], [30], [38], [39], [44]. Furthermore, some relation between the edge-Szeged index and other graph invariants were studied; see [11], [19], [24], [32]. Cai and Zhou [3] identified the n-vertex unicyclic graphs with the largest, the second largest, the smallest and the second smallest edge-Szeged indices.

In this paper, we continue the above direction of research by considering the extremal problems on the edge-Szeged index of unicyclic graphs. In the next section, we recall some known results and prove new results that are needed for the proofs of the main results. In Section 3, we determine the sharp lower bounds on the edge-Szeged index of n-vertex unicyclic graphs with diameter d. Some further discussions are given in the last section, which identify the unicyclic graphs with the minimum, second minimum, third minimum and the fourth minimum edge-Szeged indices.

Section snippets

Main results

A unicyclic graph is a (simple) connected graph with a unique cycle. The greatest distance between any two vertices in G is the diameter of G. For convenience, let Gn be the set of all n-vertex unicyclic graphs, Gnd be the set of all n-vertex unicyclic graphs with diameter d.

Let T be an n-vertex tree formed by attaching nd1 pendant vertices to ud2 of the path Pd+1=u0u1ud2ud2+1ud1ud. Then let Gnd=T+v1ud2+1, where v1 is in VTVPd+1; see Fig. 1. The following are our main results:

Theorem 2.1

For

Preliminary results

In this section, we give some preliminary results which will be used in the subsequent sections.

Fact 1

Let e=uv be a pendant edge in G, one has mu(e)mv(e)=0.

Fact 2

The function x(kx) is increasing in x for xk2.

Lemma 3.1

Let w1w2 be a cut edge of graph G and G0 be the graph obtained from G by contracting the edge w1w2 and adding a pendant edge attaching at the contracting vertex, say w1; see Fig. 2. Then Sze(G) > Sze(G0) holds if and only if w1w2 is a non-pendant edge in G.

Proof

Let G1 and G2 be the components of Gw1w2

Proof of Theorem 2.1

Recall the statement of the theorem to be proved in this section: If G is an n-vertex unicyclic graph of diameter d, then Sze(G){(d22d+8)n4d3+8d+1212,diseven;(d22d+9)n4d3+11d+2412,disoddwith equality if and only if GGnd, where Gnd is depicted in Fig. 1.

Choose G in Gnd such that Sze(G) is as small as possible. Let Pd+1=u0u1u2ud1ud be a longest path and Ck be the unique cycle in G. By Lemma 3.1, one has |VCkVPd+1|1 and each of the nontrivial pendant trees attached to the Pd+1 (or Ck) is

Proof of Theorem 2.2

In this section, based on Theorem 2.1 we use a unified approach to determine the n-vertex unicyclic graphs with the smallest, the second smallest, the third smallest and the fourth smallest edge-Szeged indices.

Lemma 5.1

If n ⩾ 6, d ⩾ 2, then Sze(Gnd)<Sze(Gnd+1).

Proof

If d is even, then d+1 is odd. In view of (4.10) we obtain Sze(Gnd+1)Sze(Gnd)=n[(d+1)22(d+1)+9]4(d+1)3+11(d+1)+2412n(d22d+8)4+d3+8d+1212d(d+2)2(d24+d2+2)=d24+d220.The inequality in (5.1) follows by the fact that nd2 is the number of

Acknowledgments

The authors would like to express their sincere gratitude to all of the referees for their criticism, insightful comments and suggestions, which led to a number of improvements to this paper.

References (47)

  • M.J. Nadjafi-Arani et al.

    On the differences between Szeged and Wiener indices of graphs

    Discr. Math.

    (2011)
  • M.J. Nadjafi-Arani et al.

    Graphs whose Szeged and Wiener numbers differ by 4 and 5, math.

    Comput. Model.

    (2012)
  • V. Nikiforov et al.

    On the minimum trace norm/energy of (0,1)-matrices

    Linear Algebra Appl.

    (2017)
  • WangS.J.

    On extremal cacti with respect to the Szeged index

    Appl. Math. Comput.

    (2017)
  • WangS.J.

    On extremal cacti with respect to the revised Szeged index

    Discr. Appl. Math.

    (2017)
  • XingR.D. et al.

    On the edge Szeged index of bridge graphs

    C. R. Math. Acad. Sci. Paris

    (2011)
  • H. Yousefi-Azari et al.

    Calculating the edge wiener and edge Szeged indices of graphs

    J. Comput. Appl. Math.

    (2011)
  • ZhangH.H. et al.

    On the further relation between the (revised) Szeged index and the wiener index of graphs

    Discr. Appl. Math.

    (2016)
  • J.A. Bondy et al.

    Graph theory

    Graduate Texts in Mathematics

    (2008)
  • CaiX.C. et al.

    Edge Szeged index of unicyclic graphs

    MATCH Commun. Math. Comput. Chem.

    (2010)
  • ChenL.L. et al.

    On a relation between the Szeged and the Wiener indices of bipartite graphs

    Trans. Comb.

    (2012)
  • A.A. Dobrynin et al.

    Wiener index of trees: theory and applications

    Acta Appl. Math.

    (2001)
  • A.A. Dobrynin et al.

    Degree distance of a graph: a degree analogue of the Wiener index

    J. Chem. Inf. Comput. Sci.

    (1994)
  • Cited by (10)

    View all citing articles on Scopus

    Financially supported by the National Natural Science Foundation of China (Grant no. 11671164).

    View full text