448
Views
1
CrossRef citations to date
0
Altmetric
Articles

-labeling of supersubdivided connected graph plus an edge

&

Abstract

Rosa, in his classical paper (Rosa, 1967) introduced a hierarchical series of labelings called ρ,σ,β and α labeling as a tool to settle Ringel’s Conjecture which states that if T is any tree with q edges then the complete graph K2q+1 can be decomposed into 2q+1 copies of T. Inspired by the result of Rosa, many researchers significantly contributed to the theory of graph decomposition using graph labeling. In this direction, in 2004, Blinco, El-Zanati and Vanden Eynden introduced γ-labeling as a stronger version of ρ-labeling. A function h defined on the vertex set of a graph G with q edges is called a γ-labeling if

(i) h is a ρ-labeling of G,

(ii) G is tripartite with vertex tripartition (A,B,C) with C={c} and b̄B such that (b̄,c) is the unique edge joining an element of B to c,

(iii) for every edge (a,v)E(G) with aA, h(a)<h(v),

(iv) h(c)h(b̄)=q.

Further, Blinco et al. proved a significant result that if a graph G with q edges admits a γ-labeling, then the complete graph K2cq+1 can be cyclically decomposed into 2cq+1 copies of the graph G, where c is any positive integer. Motivated by the result of Blinco et al., we show that a new family of almost bipartite graphs each admits γ-labeling. The new family of almost bipartite graphs is defined based on the supersubdivision graph of certain connected graph. Supersubdivision graph of a graph was introduced by Sethuraman and Selvaraju in Sethuraman and Selvaraju (2001). A graph is said to be a supersubdivision graph of a graph G with q edges, denoted SSD(G) if SSD(G) is obtained from G by replacing every edge ei of G by a complete bipartite graph K2,mi, 1iq, (where mi may vary for each edge ei) in such a way that the ends of ei are identified with the 2 vertices of the vertex part having two vertices of the complete bipartite graph of K2,mi after removing the edge ei of G. In the graph SSD(G), the vertices which originally belong to the graph G are called the base vertices of SSD(G) and all the other vertices of SSD(G) are called the non-base vertices of SSD(G). More precisely, we prove that if G is a connected graph containing a cycle Ck, where k6 and having a vertex of degree two with one of its adjacent vertices of degree one and its other adjacent vertex is of degree at least two, then certain supersubdivision graph of the graph G, SSD(G) plus an edge eˆ admits γ-labeling, where eˆ is added between a suitably chosen pair of non-base vertices of the graph SSD(G). Also, we discuss a related open problem.

1 Introduction

Decomposition of a graph H is a system R of subgraphs of H such that any edge of the graph H belongs to exactly one of the subgraphs in R. A decomposition R of a graph H is said to be cyclic if R contains a graph G then it also contains the graph G obtained by turning G. In an attempt to settle the Ringel’s Conjecture that if T is any tree with q edges then the complete graph K2q+1 can be decomposed into 2q+1 copies of T, Rosa in his classical paper [Citation1] introduced a series of labelings called ρ,σ,β and α labeling.

A one-to-one function f from the vertex set of a graph G with q edges to the set {0,1,2,,2q} is called a ρ-labeling of G if {min(|f(u)f(v)|,2q+1|f(u)f(v)|)|(u,v)E(G)}={1,2,,q}. Let G be a graph with q edges. A one-to-one function f:V(G){0,1,2,,2q} is called a σ-labeling of G if {|f(u)f(v)||(u,v)E(G)}={1,2,,q}. β-labeling of a graph G with q edges is a one-to-one function f:V(G){0,1,2,,q} such that {|f(u)f(v)||(u,v)E(G)}={1,2,,q}. β-labeling was later called as graceful labeling by Golomb [Citation2] and this term is most widely used. A β-labeling f of a graph G with q edges is called an α-labeling if there exists an integer k such that f(u)k<f(v) or f(v)k<f(u) for every edge uvE(G). It is clear that α-labeling is a stronger version of β-labeling, σ-labeling is a weaker version of β-labeling and σ-labeling is a stronger version of ρ-labeling.

Further, Rosa [Citation1] proved the following two significant theorems.

Theorem 1.1

If G is a graph with q edges, then there exists a cyclic G-decomposition of K2q+1 if and only if G has a ρ-labeling.

Theorem 1.2

Let G be a graph with q edges that has an α-labeling. Then there exists a cyclic G-decomposition of K2cq+1 into subgraphs isomorphic to G, where c is an arbitrary natural number.

The above two results inspired many researchers to discover similar labelings that can be used as a tool for decomposition of complete graphs or complete multipartite graphs. In this direction in [Citation3] El-Zanati et al. introduced ρ+-labeling. Let G be bipartite graph with q edges and bipartition (A,B). ρ+-labeling of G is a one-to-one function h:V(G){0,1,2,,2q} such that the integers h(x)h(y) are distinct modulo 2q+1 over all ordered pairs (x,y) with (x,y)E(G) and h(b)>h(a) whenever aA, bB and (a,b)E(G). They have also proved the following decomposition theorem.

Theorem 1.3

If a bipartite graph G with q edges has a ρ+-labeling and x is any positive integer then there exists a cyclic G-decomposition of K2qx+1.

In [Citation4] Fronček introduced blended ρ-labeling. Let G be a graph with 4k+1 edges, V(G)=V0V1, V0V1= and |V0|=|V1|=2k+1. Let λ be an injection, λ:Vi{0i,1i,2i,,(2k)i}, i=0,1. The pure length of an edge (xi,yi) with xi,yiVi, i{0,1} is defined as lii(xi,yi)=min{|λ(xi)λ(yi)|,2k+1|λ(xi)λ(yi)|} for i=0,1 and the mixed length of an edge (x0,y1) is defined as l01(x0,y1)=(λ(y1)λ(x0)) mod 2k+1 for x0V0,y1V1. G has a blended ρ-labeling if

(i)

{lii(xi,yi)|(xi,yi)E(G)}={1,2,,k} for i=0,1,

(ii)

{l01(x0,y1)|(x0,y1)E(G)}={1,2,,2k}.

Using the blended ρ-labeling Fronček [Citation4] proved the following decomposition theorem.

Theorem 1.4

Let the graph G with 4k+1 edges have a blended ρ-labeling. Then there exists a bi-cyclic decomposition of K4k+2 into 2k+1 copies of G.

In [Citation5] Fronček and Kubesa have examined about the decomposition of the complete graph K2n into n isomorphic spanning trees using a new type of labeling called switching blended labeling. For more details refer [Citation5]. In 2013, Anita Pasotti [Citation6] introduced a generalization of graceful labeling called d-divisible graceful labeling as a tool to obtain cyclic G-decomposition in complete m-partite graphs with parts of size n, Km×n. Let G be a graph of size e and let d be a divisor of e, say e=d.m. A d-divisible graceful labeling of G is an injective function f:V(G){0,1,2,,d(m+1)1} such that {|f(u)f(v)||(u,v)E(G)}={1,2,,d(m+1)1}{m+1,2(m+1),,(d1)(m+1)}. A d-divisible α-labeling of a bipartite graph G is a d-divisible graceful labeling of G having the property that its maximum value on one of the two bipartite sets does not reach its minimum value on the other one.

Further, Anita Pasotti [Citation6] has proved the following significant theorems.

Theorem 1.5

If there exists a d-divisible graceful labeling of a graph G of size e then there exists a cyclic G-decomposition of K(ed+1)×2d.

Theorem 1.6

If there exists a d-divisible α-labeling of a graph G of size e then there exists a cyclic G-decomposition of K(ed+1)×2dc for any positive integer c.

With the motivation to decompose the complete graph K2cq+1 into almost-bipartite graphs with q edges, where c is any positive integer, Blinco et al. [Citation7] introduced γ-labeling (A graph is said to be almost-bipartite if the removal of a particular edge makes the graph bipartite). A function h defined on the vertex set of a graph G with q edges is called a γ-labeling if

(i)

h is a ρ-labeling of G,

(ii)

G is tripartite with vertex tripartition (A,B,C) with C={c} and b̄B such that (b̄,c) is the unique edge joining an element of B to c,

(iii)

for every edge (a,v)E(G) with aA, h(a)<h(v),

(iv)

h(c)h(b̄)=n.

Further, in [Citation7], Blinco et al. have proved the following significant theorem.

Theorem 1.7

Let G be a graph with q edges having γ-labeling. Then there exists a cyclic G-decomposition of K2cq+1, where c is any positive integer.

Inspired by the above result of Blinco et al., the almost-bipartite graphs Pn+e, n4, Km,n+e, m2, n>2, C2k+1, k2, C2m+e, m>2, C3C4m, m>1, C2k+1C4n+2, k1, n1 are found to have γ-labeling (refer [Citation8–10,7,11]). For survey on γ-labeling refer the survey on graph labeling by Gallian [Citation12]. In [Citation13], Sethuraman and Selvaraju introduced a graph operation called supersubdivision of a graph that generate families of bipartite graphs from the given graph. Let G be a graph with q edges. A graph is said to be a supersubdivision graph of a graph G with q edges, denoted SSD(G) if SSD(G) is obtained from G by replacing every edge ei of G by a complete bipartite graph K2,mi, 1iq, (where mi may vary for every edge ei) in such a way that the ends of ei are identified with the 2-vertex part of K2,mi after removing the edge ei from G. (In the complete bipartite graph K2,m the part consisting of two vertices is referred as 2-vertex part of K2,m and the part consisting of m vertices is referred as m-vertex part of K2,m). Note that for 1iq, if mi=1 for any particular edge in the supersubdivision then this results in the classic definition of subdividing a single edge. Supersubdivision graph of the graph in (a) is shown in (b).

Fig. 1 (a) The graph G (b) Supersubdivision graph of the graph G.

Fig. 1 (a) The graph G (b) Supersubdivision graph of the graph G.

In [Citation14] we have proved that a family of almost bipartite graphs obtained from the supersubdivision of any tree with at least three vertices admits γ-labeling and we posed an open problem that whether supersubdivision of any connected graph plus an edge admits γ-labeling. We partially answer this question by proving the following result. Let G be a connected graph containing a cycle Ck, where k6 and having a vertex of degree two with one of its adjacent vertices of degree one and its other adjacent vertex is of degree at least two. Then certain supersubdivision graph of the graph G, SSD(G) plus an edge eˆ admits γ-labeling, where eˆ is added between a suitably chosen pair of non-base vertices of the graph SSD(G). Also, we discuss a related open problem.

2 Main result

In this section we first present Algorithm 1 to construct supersubdivision graph of a connected graph containing a cycle Ck, where k6 and having a vertex of degree two with one of its adjacent vertices of degree one and its other adjacent vertex is of degree at least two. Then we prove that certain supersubdivision graph of the graph G, SSD(G) plus an edge eˆ admits γ-labeling, where eˆ is added between a suitably chosen pair of non-base vertices of the graph SSD(G).

Algorithm 1

Construction of Supersubdivision Graphs Input. Any connected graph G containing a cycle Ck, where k6 and having a vertex of degree two with one of its adjacent vertices of degree one and its other adjacent vertex is of degree at least two.

Step 1. Naming the vertices of G by BFS algorithm.

Find a vertex of degree two having one adjacent vertex with degree one and the other adjacent vertex with degree at least two. Refer such a vertex of degree two by v, its unique adjacent vertex of degree one by w and the unique adjacent vertex of degree at least two by u. Obtain the graph G{v,w}, denote the graph thus obtained as G̃. Considering the vertex u as the root of G̃, run BFS on G̃ and obtain the BFS ordering of the vertices in G̃ as v0,v1,,vn1, where v0 is the root u and n=|V(G̃)|. Then name the vertex v as vn+1 and w as vn in the graph G. [Thus, the vertices of G are ordered (or named) as v0,v1,,vn1,vn,vn+1].

Step 3. Edge ordering

Denote the edges v0vn+1,vn+1vn as e1,e2 respectively. Among all the adjacent vertices of vn1, find the vertex with label having the largest index, say vk1. Then denote the edge vn1vk1 by e3. Then find the adjacent vertex of vn1 with the label having the largest possible index less than k1, say vk2. Then denote the edge vn1vk2 by e4. Similarly, sequentially find the adjacent vertices vk3,vk4,,vkd of vn1 and label the edges vn1vk3,vn1vk4,,vn1vkd by e5,e6,,ed+2, where d is the degree of vn1 and k1>k2>k3>kd. For j, 2jn1 in the increasing order of j, find the adjacent vertices of vnj, then label the incident edges at vnj in the increasing order of j sequentially as done above as ed+3,ed+4,,eq, where q denotes the number of edges of the graph G.

Step 4. Defining basic labels

Define f(vi)=i, for i, 0in+1.

Step 5. Edge replacement

Step 5.1. Replacement of the edge eq=v0v1

Replace the edge eq=v0v1 by K2,mq, where mq2 is any positive integer in such a way that the ends v0 and v1 of eq are identified with the 2-vertex part of K2,mq.

Step 5.2. Replacement of the edge e1=v0vn+1

Replace the edge e1=v0vn+1 by K2,m1 in such a way that the ends v0 and vn+1 of e1 are identified with the 2-vertex part of K2,m1, where m1 is defined depending on the congruence class of n+2 modulo 4 as given in .

Table 1 Value of m1 depending on the congruence class of n+2 modulo 4.

Step 5.3. Replacement of the edge e2=vn+1vn.

Replace the edge e2=vn+1vn by K2,m2 in such a way that the ends vn+1 and vn of e2 are identified with the 2-vertex part of K2,m2, where m2 is defined depending on the nature of n as given in .

Table 2 Value of m2 depending on the nature of n.

Step 5.4. Replacement of the edge ei=vxvy for i, 3iq1.

Replace every edge ei=vxvy for i, 3iq1 by K2,mi, where mi=ti|f(vx)f(vy)|=ti|xy|, 0x,yn1, ti is an arbitrary positive integer in such a way that the end vertices of ei are identified with the 2-vertex part of K2,mi.

Notation 1

For a given connected graph G, the supersubdivision graph of G constructed by Algorithm 1 is denoted by SSD(G).

Notation 2

The vertex set of the graph SSD(G), V(SSD(G)) can be partitioned into two sets B(SSD(G)) and NB(SSD(G)), where B(SSD(G)) is the set of all actual vertices of G in SSD(G) called base vertices of SSD(G) and NB(SSD(G)) is the set of vertices which lie in the mi-part of the complete bipartite graph K2,mi which replaces the edge ei of G in construction of the graph SSD(G), for 1iq and they are called non-base vertices of SSD(G).

Theorem 2.1

Let G be a connected graph containing a cycle Ck, where k6 and having a vertex of degree two with one of its adjacent vertices of degree one and its other adjacent vertex is of degree at least two. Then certain supersubdivision graph of the graph G, SSD(G) plus an edge eˆ admits γ-labeling, where eˆ is added between a suitably chosen pair of non-base vertices of the graph SSD(G).

Proof

Let G be a connected graph containing a cycle Ck, where k6 and having a vertex of degree two with one of its adjacent vertices of degree one and its other adjacent vertex is of degree at least two. Then, the supersubdivision graph of the graph G, SSD(G) is obtained by using Algorithm 1. Consider the graph SSD(G)+eˆ, where eˆ is the new edge joining any two of the non-base vertices say u and v of the mq-vertex part of the complete bipartite graph K2,mq which replaces the edge eq of the graph SSD(G). From the construction, we observe that, the graph SSD(G)+eˆ has n+2+i=1qmi vertices and M=2i=1qmi+1 edges, where n+2 is the number of vertices of the graph G, q is the number of edges of the graph G and mi is the number of vertices in the mi-part of K2,mi which replaces the edge ei in constructing SSD(G), for 1iq. That is, |V(SSD(G)+eˆ)|=n+2+i=1qmi and |E(SSD(G)+eˆ)|=M=2i=1qmi+1.

By Notation 2, V(SSD(G))=B(SSD(G))NB(SSD(G)), where B(SSD(G)) is the set of base vertices of SSD(G) and NB(SSD(G)) is the set of non-base vertices of SSD(G).

Defining vertex labeling g on the base vertices of SSD(G)+eˆ.

Define g(vi)=f(vi)=i, for i, 0in+1.

Defining vertex labeling g on the non-base vertices of SSD(G)+eˆ.

First we define labels on the non-base vertices of the complete bipartite graphs K2,m1, K2,m2 which replace the edges v0vn+1, vnvn+1 respectively in constructing SSD(G), as shown in .

Fig. 2 Labels of vertices and edges of K2,m1 and K2,m2.

Fig. 2 Labels of vertices and edges of K2,m1 and K2,m2.

For 1in2, we find all the adjacent vertices vniri1,vniri2,,vniripi of vni satisfying the index property 1ri1<ri2<<ripi<ni. [Note that, here pi denotes the number of such adjacent vertices of vni] For each i, 1in2 and for j, 1jpi, we consider the edge vnivnirij of G and the complete bipartite graph K2,tijrij which replaces the edge vnivnirijin constructing SSD(G), where tij is any positive integer and rij is obtained as |(ni)(nirij)|. Now, in order to define the labels for the non-base vertices of K2,tijrij we introduce the following parameters.

For the case j=i=1, define Nr11=n+m1+2m2+3+a, where a is defined depending on the congruence class of n+2 modulo 4 as given in .

Table 3 Value of a depending on the congruence class of n+2 modulo 4.

Define Nr1j=Nr1(j1)+2t1(j1)r1(j1), for 2jp1.

For each i, 2in2, define Nrij=Nr(i1)pi+2t(i1)pir(i1)pi1,when j=1Nri(j1)+2ti(j1)ri(j1),when 2jpi.Then, for 1in2 and for j, 1jpi, the labels of the non-base vertices of the complete bipartite graph K2,tijrij which replaces the edge vnivnirij of G in constructing the graph SSD(G) is defined as shown in . From the above labeling given in , it is clear that, except the non-base vertices of the complete bipartite graph K2,m which replaces the edge v0v1, all the other non-base vertices of the complete bipartite graph K2,tijrij which replaces the edge vnivnirij for each i, 1in2 and for j, 1jpi are labeled.

Fig. 3 Labels of vertices and edges of K2,tijrij for each i, 1in2 and for j, 1ipi.

Fig. 3 Labels of vertices and edges of K2,tijrij for each i, 1≤i≤n−2 and for j, 1≤i≤pi.

Now, we consider the complete bipartite graph K2,m which replaces the edge v0v1 in constructing SSD(G) and we define labels for the non-base vertices of K2,m as shown in .

Fig. 4 Labels of vertices and edges of the complete bipartite graph which replaces the edge v0v1.

Fig. 4 Labels of vertices and edges of the complete bipartite graph which replaces the edge v0v1.

Observation 1

Vertex labels of SSD(G) are distinct.

The labels of base vertices of SSD(G) form an increasing sequence S1:(0,1,2,,n1,n,n+1). Thus they are distinct. From , it is clear that the labels of the non-base vertices of the complete bipartite graphs K2,m1 and K2,m2 form an increasing sequence, S2:(n+2,n+5,n+7,,n+m1+2,n+m1+3,n+m1+4,n+m1+6,n+m1+8,,n+m1+2m2,n+m1+2m2+3).Note that, maximum of S1< minimum of S2. Thus, S1S2 is also an increasing sequence.

From the definition of Nr11, it is clear that the difference between the label of the first non-base vertex of the complete bipartite graph K2,t11r11 which replaces the edge vn1vn1r11 and the last non-base vertex of the complete bipartite graph K2,m2 which replaces the edge vnvn+1 is a, where the value of a is defined in .

For 2jp1, from the definition of Nr1j, it is clear that the difference between the label of the first non-base vertex of the complete bipartite graph K2,t1jrij and the label of the last non-base vertex of the complete bipartite graph K2,t1(j1)r1(j1) is Nr1jNr1(j1)(2t1(j1)1)r1(j1)+1.

For each i, 2in2 and for j, 1jpi, from the definition of Nrij, it is also clear that the difference between the label of the first non-base vertex of the complete bipartite graph K2,tijrij and the last non-base vertex of the complete bipartite graph K2,t(i1)pi1r(i1)pi1 is NrijNr(i1)pi1(2t(i1)pi11)r(i1)pi1+1.

From , we observe that the labels of the non-base vertices in the first set of K2,tijrij increase consecutively by one. Hence the labels of all the non-base vertices of the first set of K2,tijrij are distinct. Further, we observe that the least value of the labels of the first set of K2,tijrij is Nrij+rij1 and the largest value of the labels of the non-base vertices of the second set of K2,tijrij is Nrij+2rij and their difference is rij+1. As in the first set, the labels of the non-base vertices of the second set also increase by one consecutively. Hence, the labels of all the non-base vertices of the second set of K2,tijrij are also distinct. Similarly, the labels of the non-base vertices of the other remaining tij2 sets of K2,tijrij are distinct.

From , observe that the labels of the non-base vertices of the complete bipartite graph K2,m form an increasing sequence, (Nr10,Nr10+2,Nr10+4,,M1,2M1).Thus, the labels of vertices of SSD(G) can be arranged as a monotonically increasing sequence. Hence the vertex labels of the graph SSD(G) are distinct.

Observation 2

Edge labels of SSD(G)+eˆ are distinct.

Since the labels of the edges v0v,v1v are 2M1,2M2 respectively (which are beyond M), we consider their edge labels as 2M+1|g(v)g(v0)|=2M+1(2M1)=2 and 2M+1|g(v)g(v1)|=2M+1(2M2)=3 respectively. From , we observe that the labels of the edges of K2,m1 and K2,m2 can be arranged as the following sequences, 1,(4,5,6,,m1+1,m1+2),(m1+3,m1+4,m1+5,,m1+2m21,m1+2m2),(n+2=m1+2m2+1),(m1+2m2+2,m1+2m2+3),(n+5=m1+2m2+4,n+6=m1+2m2+5,,n+m1+2=2m1+2m2+1,n+m1+3=2m1+2m2+2).From , we observe that the 2rij edges of the first set of K2,tijrij get distinct values from Nrijn+i to Nrijn+i+2rij1,the 2rij edges of the second set of K2,tijrij get distinct values from Nrijn+i+2rij to Nrijn+i+4rij1and finally, the 2rij edges of the tijth set of K2,tijrij get distinct values from Nrijn+i+2(tij1)rij to Nrijn+i+2tijrij1.Thus, the 2tijrij edges of K2,tijrij can be arranged as a sequence, Nrijn+i,Nrijn+i+1,,Nrijn+i+2tijrij2,Nrijn+i+2tijrij1.From , the labels of the edges of K2,m can be arranged as three sequences, (Nr101,Nr10,Nr10+1,Nr10+2,,M2,M1),M,(3,2).Thus, from , it is clear that the labels of edges of SSD(G)+eˆ can be arranged as a monotonically increasing sequence from 1 to M.

Hence the edge labels of SSD(G)+eˆ are distinct.

Observation 3

g is a γ-labeling.

In order to prove that g is a γ-labeling, we partition the vertex set V(SSD(G)) as (X,Y,Z), where X=B(SSD(G)), Y=NB(SSD(G)){v} and Z={v}. Then, by the above labeling, we have g(vk)<g(uij) for any vkX and for any uijYZ.

The label of the edge uv=M=(2M1(M1)).

Hence, from Observations 1–3, the graph SSD(G)+eˆ admits γ-labeling. □

Illustration

We illustrate below the γ-labeling that is defined as in the proof of Theorem 2.1. The connected graph G with edge labels is given in .

Fig. 5 The connected graph G with edge labels.

Fig. 5 The connected graph G with edge labels.

The γ-labeled SSD(G)+eˆ, where the γ-labeling as defined in the proof of Theorem 2.1 is given in . Note that eˆ=(104,209).

Fig. 6 γ-labeling of SSD(G)+eˆ.

Fig. 6 γ-labeling of SSD(G)+eˆ.

As the graph SSD(G)+eˆ admits γ-labeling, from Theorems 1.7 and 2.1 we have the following corollary.

Corollary 2.2

The complete graph K2cm+1 can be cyclically decomposed into copies of the graph SSD(G)+eˆ, where G is a connected graph containing a cycle Ck, where k6 and having a vertex of degree two with one of its adjacent vertices of degree one and its other adjacent vertex is of degree at least two, SSD(G)+eˆ is certain supersubdivision graph of the graph G plus an edge eˆ added between suitably chosen pair of non-base vertices of the graph SSD(G), c is any positive integer and m=|E(SSD(G)+eˆ)|.

3 Discussion

In Section 2, we consider the connected graph G containing a cycle Ck, where k6 and having a vertex of degree two with one of its adjacent vertices of degree one and its other adjacent vertex is of degree at least two. Then the graph SSD(G)+eˆ is obtained from certain supersubdivision graph of the graph G and adding an edge eˆ between suitably chosen pair of non-base vertices in SSD(G). In Theorem 2.1 for such special connected graph G, we have shown that SSD(G)+eˆ admits γ-labeling. We strongly feel that supersubdivision of any connected graph G with one additional edge would also admit γ-labeling and also from the inspiration of the result of Sethuraman and Selvaraju [Citation15] we pose the following conjecture,

Supersubdivision of any connected graph G plus an edge eˆ, SSD(G)+eˆ admits γ-labeling, where eˆ is added between suitably chosen pair of non-base vertices of SSD(G).

References

  • RosaA., On certain valuations of the vertices of a graph Theory of Graphs Rome, 1966 International Symposium1967Gordon and Breach, NY, Dunod Paris349–355
  • GolombS.W., How to number a graphReadR.C.Graph Theory and computing1972Academic PressNew York23–37
  • El-ZanatiS.I.Vanden EyndenC.PuninN., On the cyclic decomposition of complete graphs into bipartite graphs Australas. J. Combin. 242001 209–219
  • FrončekD., Bi-cyclic decompositions of complete graphs into spanning trees Discrete Math. 3072007 1317–1322
  • FrončekD.KubesaM., Factorizations of complete graphs into spanning trees Congr. Numer. 1542002 125–134
  • PasottiA., On d-graceful labelings Ars Combin. 1112013 207–223
  • BlincoA.El-ZanatiS.I.Vanden EyndenC., On the cyclic decomposition of complete graphs into almost-bipartite graphs Discrete Math. 2842004 71–81
  • BlairG.W.BowmanD.L.El-ZanatiS.I.HaldS.M.PribanM.K.SebestaK.A., On cyclic C2m+e-designs Ars Combin. 932009 289–304
  • BungeR.C.El-ZanatiS.I.O’HanlonW.Vanden EyndenC., On γ-labeling of the almost-bipartite graph Pm+e Ars Combin. 1072012 65–80
  • El-ZanatiS.I.O’HanlonW.A.SpicerE.R., On γ-labeling of the almost-bipartite graph Km,n+e East-West J. Math. 10 2 2008 133–139
  • El-ZanatiS.I.Vanden EyndenC., On Rosa-type labelings and cyclic graph decompositions Math. Slovaca 59 1 2009 1–18
  • GallianJ.A., A dynamic survey of graph labeling Electron. J. Combin. 202017 #DS6
  • SethuramanG.SelvarajuP., Gracefuness of arbitrary supersubdivisions of graphs Indian J. Pure Appl. Math. 32 7 2001 1059–1064
  • G. Sethuraman, M. Sujasree, Generating γ-labeled graphs from any tree with at least three vertices, Indian J. Pure Appl. Math. (submitted for publication).
  • SethuramanG.SelvarajuP., Decompositions of complete graphs and complete bipartite graphs into isomorphic supersubdivision graphs Discrete Math. 2602003 137–149