International Journal on Recent and Innovation Trends in Computing and Communication ISSN: 2321-8169

Volume: 5 Issue: 6 293 304

_______________________________________________________________________________________________

The Sum-Eccentricity Energy Of A Graph

Mohammad Issa Sowaity and B. Sharada

Department of Studies in Mathematics, University of Mysore, Man-

asagangotri, Mysuru - 570 006, I NDIA
E-mail address: mohammad_d2007@hotmail.com

Department of Studies in Computer Scince, University of Mysore ,

Manasagangotri , Mysuru - 570 006, I NDIA
E-mail address: sharadab21@gmail.com

A b st r ac t . In this paper, we introduce the concept of the sum -eccentricity

matrix S e (G) of a graph G and obtain some coefficients of the characteristic
polynomial P(G, ) of the sum-eccentricity matrix of G . We also introduce the
sum-eccentricity energy ES e (G) of a graph G . Sum-eccentricity energies of
some well-known graphs are obtained. Upper and lower bounds for ES e (G) are
estblished. It is shown that if the sum -eccentricity energy of a graph is
rational then it must be an even.

Key words and phrases. Distance in graphs, Sum-eccentricity matrix, Sum-

eccentricity eigenvalues, Sum-eccentricity energy of a graph.

_____________________________________________*****___________________________________________

1. Introduction
In this paper, all graphs are assumed to be finite connected simple graphs.
A graph G (V , E ) is a simple graph, that is, having no loops, no multiple
and directed edges. As usual, we denote n to be the order and m to
be the size of the graph G . For a vertex v V , the open neighborhood
of v in a graph G , denoted N (v), is the set of all vertecies that are
adjacent to v and the closed neighborhood of v is N[v] N (v) {v}. The
degree of a vertex v in G is d (v) | N (v) | . The distance d (u, v) between
any two vertices u and v in a graph G is the length of the shortest
293
IJRITCC | June 2017, Available @ http://www.ijritcc.org
_______________________________________________________________________________________
International Journal on Recent and Innovation Trends in Computing and Communication ISSN: 2321-8169
Volume: 5 Issue: 6 293 304
_______________________________________________________________________________________________
path connecting them. The eccevtricity of a vertex v G is
e(v) max{ d (u, v):u V (G)}. The radius of G is r (G) min{e(v):v V (G)} and the
diameter of G is D(G) max{ e(v):v V (G)}. Hence r (G) e(v)( D(G), for
every v V (G). A vertex v in a connected graph G is central if
e(v) r (G), while a vertex v in a connected graph G is peripheral vertex
if e(v) D(G). A graph G is called self centered graph if e(v) r (G) D(G).
The girth of a graph G is the length of the shortest cycle contained in
the graph and denoted by g (G). All the defnitions and terminologies
about the graph in this paragraph available in [9].

The concept energy of a graph introduced by I. Gutman [8], in (1978). Let

G be a graph with n vertices and m edges and let A(G) (aij ) be the
adjacency matrix of G, where
1, if vi v j E ,
aij
0 , otherwise.

The eigenvalues 1 , 2 ,, n of a matrix A(G) assumed in a non-increasing

order, are the eigenvalues of a graph G [10]. Let 1 2 t , for t n
be the distinct eigenvalues of G with multiplicities m1 , m2 ,, mt ,
respectively, the multiset of eigenvalues of A(G) is called the spectrum
of G and denoted by
2 t
Sp(G) 1
m1 m2 mt
As A is real symmetric with zero trace, the eigenvalues of G are real
with sum equal to zero [3]. The energy E (G) of a graph G is defined to
be the sum of the absolute values of the eigenvalues of G [8], i.e.,
n
E (G ) | i |.
i 1

For more details on the mathematical aspects of the theory of graph

energy we refer to [5, 7, 10] and the references therein.

C. Adiga et. al. [2], have defined the maximum degree energy EM (G) of a
graph G which depends on the maximum degree matrix M (G) of G. Let
G be a simple graph with n vertices v1 , v2 ,, vn . Then the maximum
degree matrix M (G) (d ij ) of a graph G defined as

294
IJRITCC | June 2017, Available @ http://www.ijritcc.org
_______________________________________________________________________________________
International Journal on Recent and Innovation Trends in Computing and Communication ISSN: 2321-8169
Volume: 5 Issue: 6 293 304
_______________________________________________________________________________________________
max{ d (vi ), d (v j )}, if vi v j E,
d ij
0 , otherwise.
As M (G) is real symmetric with zero trace, then the eigenvalues of G
being real with sm equal to zero.

Ahmed M. Naji et. al. [3], have defined the concept of maximum
eccentricity matrix M e (G) of a connected graph G. They obtained the
maximum eccentricity energy EM e (G) of a graph depends on the
maximum eccentricity matrix. Let G be a simple connected graph with n
vertices v1 , v2 ,, vn and let e(vi ) be the eccentricity of a vetex vi , i 1, 2,..., n
The maximum eccentricity matrix of G defined as M e (G) (eij ), where
max{ e(vi ), e(v j )}, if vi v j E,
eij
0 , otherwise.
Motivated by those papers, we introduce the concept of the sum -
eccentricity matrix S e (G) of a graph G and obtain some coefficients of
the characteristic polynomial P(G, ) of the sum-eccentricity matrix of G .
We also introduce the sum-eccentricity energy ES e (G) of a graph G.
Sum-eccentricity energies of some well -known graphs are obtained.
Upper and lower bounds for ES e (G) are estblished. It is shown that if
the sum-eccentricity energy of a graph is rational then it must be an
even.

2. THE SUM-ECCENTRICITY ENERGY OF GRAPHS

Definition 2.1. Let G be a graph with n vertices. Then the sum-

eccentricty matrix of a graph G denoted by S e (G), is defined as
S e (G) (sij ), where
e(vi ) e(v j ) , if vi v j E ,
sij
0 , otherwise.
The characteristic polynomial of the sum -eccentricity matrix S e (G) is
defined by
P(G, ) det(I Se (G)),
Where I is the unt matrix of order n The eigenvalues of the sum-
eccentricity matrix S e (G) are the roots of the charecteristic polynomial of

295
IJRITCC | June 2017, Available @ http://www.ijritcc.org
_______________________________________________________________________________________
International Journal on Recent and Innovation Trends in Computing and Communication ISSN: 2321-8169
Volume: 5 Issue: 6 293 304
_______________________________________________________________________________________________
G.
Since S e (G) is real symmetric with zero trace, its eigenvalues must be
realwith sum equal to zero, i.e., trace(Se (G)) 0. We lable the eigenvalues
1 , 2 ,, n in a non-increasing manner 1 2 n . The sum-eccentrcity
energy of a graph G is denoted by ES e (G) and is defined as the
summation of the absolute value of the eigenvalues
n
ES e (G) | i | .
i 1

The following examples explain the concept.

Example 2.2.Let G1 be the graph as in figure 1.

Then the sum-eccentricity matrix of G1 is

0 6 0 0 0 5
6 0 5 00 5

0 5 0 0 4 0
S e (G1 )
0 0 4 5 0 4
5 5 0 4 0 0

0 0 0 5 0 0
The charracteristic polynomial of S e (G1 ) is
P(G1 , ) 6 1684 3003 49522 7500 15625.
The sum-eccentricity eigenvalues of G1 are
1 12.54, 2 5.4884, 3 1.2211, 4 2.8779, 5 6.6336, 6 9.7383.
The sum-eccentricity energy of G1 is
ES e (G1 ) 38.499.
Example 2.3. Let G2 be the K 5 graph.

Then the sum-eccentricity matrix of G2 is

0 2 2 2 2
2 0 2 2 2

S e (G2 ) 2 2 0 2 2

2 2 2 0 2
2 2 2 2 0

296
IJRITCC | June 2017, Available @ http://www.ijritcc.org
_______________________________________________________________________________________
International Journal on Recent and Innovation Trends in Computing and Communication ISSN: 2321-8169
Volume: 5 Issue: 6 293 304
_______________________________________________________________________________________________
The charracteristic polynomial of S e (G2 ) is
P(G2 , ) 5 403 162 240 128 ( 2) 4 ( 8).
The sum-eccentricity eigenvalues of G2 are
1 8, 2 2, 3 2, 4 2, 5 2.
The sum-eccentricity energy of G2 is
ES e (G2 ) 16.
3. BOUNDS FOR SUM-ECCENTRICITY ENERGY AND SUM-
ECCENTRICITY EIGENVALUES
We now give the explicit expression for the coefficient ci of
ni (i 0,1, 2, 3 and n) in the characteristic polynomial of the sum -eccentricity
matrix S e (G).

Theorem 3.1. Let G be a graph of order n and let

P(G, ) c0 n c1n1 c2 n2 ... cn ,
be the charracteristic polynomial of S e (G). Then
1. c0 1.
2. c1 0.
n
3. c2 (e(v ) e(v ))
i 1, i j
i j
2
, where vi v j E.
n
4. c3 2 (2e(v )e(v )e(v
vi v j vk ,1 i j k n
i j k ) e(vi ) 2 e(v j ) e(vi ) 2 e(vk ) e(v j ) 2 e(vi )

e(v j ) 2 e(vk ) e(vk ) 2 e(vi ) e(vk ) 2 e(v j )).

5. For n 1we have cn (1) n det (S e (G)).

Proof. The proof of parts (1) and (2) are similar to the proof in [2].
3. Since
0 sij
c2 0 (s s ) s
2
ij ji ij
1i j n s ji 0 1i j n 1i j n

and since
e(vi ) e(v j ) , if vi v j E ,
sij
0 , otherwise.
n
Thus c2 (e(v ) e(v ))
i 1, i j
i j
2
, where vi v j E.

297
IJRITCC | June 2017, Available @ http://www.ijritcc.org
_______________________________________________________________________________________
International Journal on Recent and Innovation Trends in Computing and Communication ISSN: 2321-8169
Volume: 5 Issue: 6 293 304
_______________________________________________________________________________________________

4. We have
sii n
sij sik
c2 s ji s jj s jk
1i j k n
ski skj skk
n
2 (s s
1i j k n
ij ik s jk )
n
2 [(e(v ) e(v ))(e(v ) e(v
vi v j vk ,1 i j k n
i j i k ))(e(v j )e(vk ))]

n
2 (2e(v )e(v )e(v
vi v j vk ,1 i j k n
i j k ) e(vi ) 2 e(v j ) e(vi ) 2 e(vk ) e(v j ) 2 e(vi )

e(v j ) 2 e(vk ) e(vk ) 2 e(vi ) e(vk ) 2 e(v j )).

n
5. We have ck (1) (all k k principle min ors)
k

k 1

hence cn (1) det (S e (G)).

n

Example 3.2.In the graph G1 in figure 1, the coefficient c2 of 4 in the

characteristic polynomialof S e (G1 ) is equal to
n
(e(v ) e(v ))
i 1, i j
i j
2
, where vi v j E

[(3 3) 2 (3 2) 2 (3 2) 2 (3 2) 2 (2 2) 2 (2 2) 2 (2 3) 2 ] 168

Remark 3.3. a.The number of terms in c3 in the above theorem is equal to

the number of triangles in the graph.

b. If g (G) 3, then c3 0.

Theorem 3.4.If 1 , 2 ,, n , are the sum-eccentricity eigenvalues of a graph

G, then

i 1
i
2
2c2 .

Proof. We have

n n n n n n

i 2 trace(Se2 (G))
i 1 i 1
sik ski 2
k 1 i 1
sik2 2
ik
s
i 1, i k
2
ik

298
IJRITCC | June 2017, Available @ http://www.ijritcc.org
_______________________________________________________________________________________
International Journal on Recent and Innovation Trends in Computing and Communication ISSN: 2321-8169
Volume: 5 Issue: 6 293 304
_______________________________________________________________________________________________
n
2 (e(v ) e(v ))
i 1, i k
i k
2
, where vi vk E ,

hence

i 1
i
2
2c2 .

Theorem 3.5. Let G K n , a complete graph of order n, n 1, then

c2 2n(n 1).

n
Proof. We have c2 (e(v ) e(v ))
i 1, i j
i j
2
, where vi v j E,

we also have in K n each e(vi ) 1 so

n 1
n(n 1)
c2 (2 2) 2 i 4 2n(n 1).
i 1 2

Example 3.6. In the graph G2 , the coefficient c2 of 3 in the

characteristic polynomialof S e (G2 ) is 2(5)(4) 40.

Corollary 3.7. For the complete graph K n , we have

i 1
i
2
4n(n 1).

Theorem 3.8. If G is a graph of order n, then for any sum-

eccentricity eigenvalue j , we have

(n 2)2j
c2 2n((n 1) 2 .
2

Proof. We have

trace(Se2 ( K n )) 4n(n 1)
299
IJRITCC | June 2017, Available @ http://www.ijritcc.org
_______________________________________________________________________________________
International Journal on Recent and Innovation Trends in Computing and Communication ISSN: 2321-8169
Volume: 5 Issue: 6 293 304
_______________________________________________________________________________________________
by Cauchy-Schwartz inequality, we have

n n

i2 (n 1)
i 1, i j

i 1, i j
2
i (n 1)(4n(n 1) 2j )

so

i 1, i j
2
i 4n(n 1) 2 2j (n 1)

n
i.e. i2 4n(n 1) 2 2j (n 1) 2j 4n(n 1) 2 2j (n 2).
i 1

Using theorem 3.4., we get

(n 2)2j
c2 2n((n 1) 2 .
2

Theorem 3.9. We have

2n 2 c2 4n 3 (n 1) 2
n 2
2 (e(vi ) e(v j )) 2 n(n 1) Ln ES e (G)
i 1, i j n2
,

n
where vi v j E , L i and n 2 for the left side of the inequality.
i 1

Proof. We have

n
E 2 S e (G) ( | i |) 2
i 1

n
| i |2 | i || j |.
i 1 i j

Using the last inequality in theorem 3.1 and Arithmatic mean, Geometric
mean inequality we get

n
E S e (G) 2
2
(e(v ) e(v )) |
i 1, i j
i j
2

i j
i || j |, where vi v j E,

but

300
IJRITCC | June 2017, Available @ http://www.ijritcc.org
_______________________________________________________________________________________
International Journal on Recent and Innovation Trends in Computing and Communication ISSN: 2321-8169
Volume: 5 Issue: 6 293 304
_______________________________________________________________________________________________

|
i j
i || j | | 1 | (| 2 | | 3 | ... | n |)

| 2 | (| 1 | | 3 | ... | n |)

| n | (| 1 | | 2 | ... | n1 |)

1 1

n(n 1)(| 1 || 2 || 3 | ... | n |) (| 1 | | 2 | | 3 |

n n1 n1 n1
... | n | )
n1 n ( n1)

hence

n 2
2 (e(vi ) e(v j )) 2 n(n 1) Ln ES e (G),
i 1, i j

n
where vi v j E and L i .
i 1

On the other hand, using the previous theorem we have

2c2 4n(n 1) 2
| j | ,
n2

so

n
2n 2 c2 4n 3 (n 1) 2
| j |
j 1 n2
, where n 2.

Theorem 3.10. If the sum-eccentricity energy of a graph G is

retional, then it must be an even integer.

Proof. Let 1 , 2 ,, n be the sum-eccentricity eigenvalues of a graph G

n
with order n . Then we have
i 1
i 0. Let 1 , 2 ,, r be positive, and

r 1 , r 2 ,, n arenon-positive. Then,
301
IJRITCC | June 2017, Available @ http://www.ijritcc.org
_______________________________________________________________________________________
International Journal on Recent and Innovation Trends in Computing and Communication ISSN: 2321-8169
Volume: 5 Issue: 6 293 304
_______________________________________________________________________________________________
ES e (G) 2(1 2 r ).

Since 1 , 2 ,, r are algebraic numbers, so is there sum, and hence

must be integer if ES e (G) is retional. Thus ES e (G) is an even positive

integer if it is rational.

4. THE SUM-ECCENTRICITY ENERGY FOR SOME STANDARD

GRAPHS
In this section we investigate the exact values of the sum -eccentricity
energy of some well-known graphs.
Theorem 4.1. For the cycle Cn , n 3, is we have

n 3 , if n is even,
c2
n(n 1) 2 , if n is odd .

n
Proof. We have c2 (e(v ) e(v ))
i 1, i j
i j
2
,

and

n
2 , if n is even,
e(vi )
(n 1) , if n is odd ,
2

n
n
so if n is even c2 (2 2 )
i 1, i j
2
n 3 ,

n
n 1 2
and if n is odd c2 (2
i 1, i j 2
) n(n 1) 2 ,

thus

n 3 , if n is even,
c2
n(n 1) 2 , if n is odd .

302
IJRITCC | June 2017, Available @ http://www.ijritcc.org
_______________________________________________________________________________________
International Journal on Recent and Innovation Trends in Computing and Communication ISSN: 2321-8169
Volume: 5 Issue: 6 293 304
_______________________________________________________________________________________________
Theorem 4.1. The sum-eccentricity eigenvalues for the complete graph
K n are 2 and 2(n 1) with multiplicities (n 1) and 1 respectively, and the sum-
eccentricity energy for K n is 4(n 1).

Proof. We have

2 2 2
2 2 2
| I S e ( K n ) | 2 2 2

2 2 2

2 2 2
1 1 0 0
( 2) n 1
1 0 1 0

1 0 0 1

( 2) n1 ( 2(n 1)).

The sum-eccentricity eigenvalues of K n are 1 2(n 1), 2 2, 3 2,..., n 2,

i.e., 2 with multiplicity n 1 and 2(n 1) with multiplicity 1.

Hence ES e ( K n ) 4(n 1).

References

[1] C. Adiga, E. Sampathkumar, M. A. Sriraj and Shrikanth A. S., Color energy

of a graph, Jangjeon Math. Society, 16(2013), No. 3, 335-351.
[2] C. Adiga and M. Smitha, On maximum degree energy of a graph, Int. J.
Contempt. Math. Sci., 4(2009), 385-396.
[3] M. Ahmed Naji and N. D. Soner, The maximum eccentricity energy of a graph, Int.
J. Sci. Engin. Research, 7(2016), 5-13.
[4] R. Balakrishnan, The energy of a graph, Linear Alg. Appl., 387(2004), 287-295.
[5] R. B. Bapat, Graphs and Matricies, Hindustan Book Agency, 2011.
[6] D. Cherny, T. Denton, R. Thomas and Waldron, Linear Algebra, Edited by Katrina
303
IJRITCC | June 2017, Available @ http://www.ijritcc.org
_______________________________________________________________________________________
International Journal on Recent and Innovation Trends in Computing and Communication ISSN: 2321-8169
Volume: 5 Issue: 6 293 304
_______________________________________________________________________________________________
Glaeser and Travis Scrimshaw First Edtion. Davis California, 2013.
[7] I. Gutman, X. Li and J. Zhang, Graph energy, (Ed-s: M. Dehmer, F. Emmert),
Streib. Analysis of Complex Networks, From Biology to Linguistics, Wiley-VCH,
Weinheim, (2009), 154-174.
[8] I. Gutman, The energy of a graph, Ber. Math. Stat. Sekt. Forsch. Graz, 103(1978),
1-22.
[9] F. Harary, Graph Theory, Addision-Wesley Publishing Co., Reading, Mass. Menlo
park, Calif. London, 1969.
[10] X. Li, Y. Shi and I. Gutman, Graph Energy, Springer, 2012.

304
IJRITCC | June 2017, Available @ http://www.ijritcc.org
_______________________________________________________________________________________