arXiv:1003.4855v2 [math.

CO] 28 Jul 2010

A note on the partition dimension of Cartesian

product graphs
Ismael G. Yero and Juan A. Rodrguez-Velazquez
Departament dEnginyeria Inform`atica i Matem`atiques
Universitat Rovira i Virgili, Av. Pasos Catalans 26, 43007 Tarragona, Spain.
ismael.gonzalez@urv.cat, juanalberto.rodriguez@urv.cat

July 29, 2010

Abstract
Let G = (V, E) be a connected graph. The distance between two
vertices u, v V , denoted by d(u, v), is the length of a shortest u v
path in G. The distance between a vertex v V and a subset P V
is defined as min{d(v, x) : x P }, and it is denoted by d(v, P ). An
ordered partition {P1 , P2 , ..., Pt } of vertices of a graph G, is a resolving
partition of G, if all the distance vectors (d(v, P1 ), d(v, P2 ), ..., d(v, Pt ))
are different. The partition dimension of G, denoted by pd(G), is the
minimum number of sets in any resolving partition of G. In this
article we study the partition dimension of Cartesian product graphs.
More precisely, we show that for all pairs of connected graphs G, H,
pd(G H) pd(G) + pd(H) and pd(G H) pd(G) + dim(H), where
dim(H) denotes the metric dimension of H. Consequently, we show
that pd(G H) dim(G) + dim(H) + 1.

Keywords: Resolving sets, resolving partition, partition dimension, Cartesian product.

AMS Subject Classification numbers: 05C12; 05C70; 05C76

Introduction

The concepts of resolvability and location in graphs were described independently by Harary and Melter [9] and Slater [17], to define the same structure
in a graph. After these papers were published several authors developed diverse theoretical works about this topic [2, 3, 4, 5, 6, 7, 8, 14]. Slater described
the usefulness of these ideas into long range aids to navigation [17]. Also,
these concepts have some applications in chemistry for representing chemical
compounds [12, 13] or to problems of pattern recognition and image processing, some of which involve the use of hierarchical data structures [15].
Other applications of this concept to navigation of robots in networks and
other areas appear in [5, 11, 14]. Some variations on resolvability or location
have been appearing in the literature, like those about conditional resolvability [16], locating domination [10], resolving domination [1] and resolving
partitions [4, 7, 8].
Given a graph G = (V, E) and an ordered set of vertices S = {v1 , v2 , ..., vk }
of G, the metric representation of a vertex v V with respect to S is the
vector r(v|S) = (d(v, v1 ), d(v, v2), ..., d(v, vk )), where d(v, vi ), with 1 i k,
denotes the distance between the vertices v and vi . We say that S is a resolving set of G if for every pair of distinct vertices u, v V , r(u|S) 6= r(v|S).
The metric dimension1 of G is the minimum cardinality of any resolving set
of G, and it is denoted by dim(G). The metric dimension of graphs is studied
in [2, 3, 4, 5, 6, 18].
Given an ordered partition = {P1 , P2 , ..., Pt } of the vertices of G, the
partition representation of a vertex v V with respect to the partition
is the vector r(v|) = (d(v, P1), d(v, P2), ..., d(v, Pt )), where d(v, Pi ), with
1 i t, represents the distance between the vertex v and the set Pi , that
is d(v, Pi ) = minuPi {d(v, u)}. We say that is a resolving partition of G if
for every pair of distinct vertices u, v V , r(u|) 6= r(v|). The partition
dimension of G is the minimum number of sets in any resolving partition of
G and it is denoted by pd(G). The partition dimension of graphs is studied
in [4, 7, 8, 18]. It is natural to think that the partition dimension and metric
dimension are related; in [7] it was shown that for any nontrivial connected
graph G we have
pd(G) dim(G) + 1.
(1)
The study of relationships between invariants of Cartesian product graphs
1

Also called locating number.

and invariants of its factors appears frequently in research about graph theory. In the case of resolvability, the relationships between the metric dimension of the Cartesian product graphs and the metric dimension of its
factors was studied in [2, 3]. An open problem on the dimension of Cartesian
product graphs is to prove (or finding a counterexample) that for all pairs
of graphs G, H; dim(G H) dim(G) + dim(H). In the present paper we
study the case of resolving partition in Cartesian product graphs, by giving some relationships between the partition dimension of Cartesian product
graphs and the partition dimension of its factors. More precisely, we show
that for all pairs of connected graphs G, H; pd(G H) pd(G) + pd(H) and
pd(G H) pd(G) + dim(H). Consequently, we show that pd(G H)
dim(G) + dim(H) + 1.
We recall that the Cartesian product of two graphs G1 = (V1 , E1 ) and
G2 = (V2 , E2 ) is the graph G1 G2 = (V, E), such that V = {(a, b) : a
V1 , b V2 } and two vertices (a, b) V and (c, d) V are adjacent in G1 G2
if and only if, either a = c and bd E2 or b = d and ac E1 .
The following well known fact will be used several times.
Remark 1. Let the graph Gi = (Vi , Ei ) and let Si Vi , i {1, 2}. For every
(a, b) V1 V2 , it follows dG1 G2 ((a, b), S1 S2 ) = dG1 (a, S1 ) + dG2 (b, S2 ).

The partition dimension of Cartesian product graphs

Theorem 2. For any connected graphs G1 and G2 ,

pd(G1 G2 ) pd(G1 ) + pd(G2 ).
Proof. Let 1 = {A1 , A2 , ..., Ak } and 2 = {B1 , B2 , ..., Bt } be resolving partitions of G1 = (V1 , E1 ) and G2 = (V2 , E2 ) respectively. Let us show that
= {A1 B1 , A1 B2 , ..., A1 Bt , A2 B1 , A3 B1 , ..., Ak B1 , C}, with
C = (V1 V2 ) ((V1 B1 ) (A1 V2 )) is a resolving partition of G1 G2 .
Let (a, b), (c, d) be two different vertices of V1 V2 . If a = c, then there
exists Bi 2 such that dG2 (b, Bi ) 6= dG2 (d, Bi ). Hence we have
dG1 G2 ((a, b), A1 Bi ) =dG1 (a, A1 ) + dG2 (b, Bi )
6=dG1 (c, A1 ) + dG2 (d, Bi )
=dG1 G2 ((c, d), A1 Bi )
3

Now, if a 6= c then we have the following cases:

Case 1: Let a Ai and c Aj , with i 6= j. If we suppose,
dG1 G2 ((a, b), Ai B1 ) = dG1 G2 ((c, d), Ai B1 )
and
dG1 G2 ((a, b), Aj B1 ) = dG1 G2 ((c, d), Aj B1 ),
we obtain
dG2 (b, B1 ) =dG1 G2 ((a, b), Ai B1 )
=dG1 G2 ((c, d), Ai B1 )
=dG1 (c, Ai ) + dG2 (d, B1)
=dG1 (c, Ai ) + dG1 G2 ((c, d), Aj B1 )
=dG1 (c, Ai ) + dG1 G2 ((a, b), Aj B1 )
=dG1 (c, Ai ) + dG1 (a, Aj ) + dG2 (b, B1 ),
a contradiction.
Case 2: If a, c Ai then we have the following subcases.
Case 2.1: b, d Bl . Let Aj 1 , such that dG1 (a, Aj ) 6= dG1 (c, Aj ). In
this case, if dG2 (b, B1 ) = dG2 (d, B1 ) then we have
dG1 G2 ((a, b), Aj B1 ) =dG1 (a, Aj ) + dG2 (b, B1 )
6=dG1 (c, Aj ) + dG2 (d, B1 )
=dG1 G2 ((c, d), Aj B1 ).
On the contrary, if dG2 (b, B1 ) 6= dG2 (d, B1) then we have
dG1 G2 ((a, b), Ai B1 ) =dG2 (b, B1 )
6=dG2 (d, B1 )
=dG1 G2 ((c, d), Ai B1 ).
Case 2.2: b Bj and d Bl , j 6= l. This case is analogous to Case 1.
Therefore for every different vertices (a, b), (c, d) V1 V2 , we have
r((a, b)|) 6= r((c, d)|).
By (1) we obtain the following direct consequence of Theorem 2.
Corollary 3. For any connected graphs G1 and G2 ,
pd(G1 G2 ) pd(G1 ) + dim(G2 ) + 1.
4

As we can see below, the above relationship can be improved.

Theorem 4. For any connected graphs G1 and G2 ,
pd(G1 G2 ) pd(G1 ) + dim(G2 ).
Proof. Let = {A1 , A2 , ..., Ak } be a resolving partition of G1 = (V1 , E1 ),
let S = {u1 , u2, ..., ut } be a resolving set of G2 = (V2 , E2 ) and let C =
V1 V2 ((V1 {u1 }) (A1 {u2 }) (A1 {ut })). Let us show that
1 = {A1 {u1 }, A2 {u1 }, ..., Ak {u1 }, A1 {u2 }, A1 {u3 }, ..., A1 {ut }, C}
is a resolving partition of G1 G2 .
Let (a, b), (c, d) be two different vertices of V1 V2 . If a = c, then b 6= d.
Thus, there exist uj S such that dG2 (b, uj ) 6= dG2 (d, uj ). Hence,
dG1 G2 ((a, b), A1 {uj }) =dG1 (a, A1 ) + dG2 (b, uj )
6=dG1 (c, A1 ) + dG2 (d, uj )
=dG1 G2 ((c, d), A1 {uj })
Now, if a 6= c we have two cases:
Case 1: a Ai and c Aj , j 6= i. Let us suppose, dG2 (b, u1 ) dG2 (d, u1 ).
In this case we have
dG1 G2 ((a, b), Ai {u1 }) =dG2 (b, u1 )
dG2 (d, u1 )
<dG1 (c, Ai ) + dG2 (d, u1)
=dG1 G2 ((c, d), Ai {u1 }).
Analogously, if dG2 (b, u1 ) dG2 (d, u1) we obtain
dG1 G2 ((a, b), Aj {u1 }) > dG1 G2 ((c, d), Aj {u1 }).
Case 2: a, c Ai . Let us suppose dG2 (b, u1 ) = dG2 (d, u1). Since there exists
j 6= i, such that dG (a, Aj ) 6= dG (c, Aj ), we have
dG1 G2 ((a, b), Aj {u1 }) =dG1 (a, Aj ) + dG2 (b, u1 )
6=dG1 (c, Aj ) + dG2 (d, u1)
=dG1 G2 ((c, d), Aj {u1 }).
If dG2 (b, u1 ) 6= dG2 (d, u1), we have dG1 G2 ((a, b), Ai {u1 }) = dG2 (b, u1 ) 6=
dG2 (d, u1) = dG1 G2 ((c, d), Ai {u1 }). Therefore, for every different vertices
(a, b), (c, d) we have r((a, b)|1 ) 6= r((c, d)|1).
5

In order to give some examples we emphasize the following well known

values for the metric dimension of the complete graph, Kn , the path graph,
Pn , the cycle graph, Cn , and the star graph, K1,n .
Remark 5.
(i) dim(Kn ) = n 1 (n 2).
(ii) dim(Pn ) = 1.
(iii) dim(Cn ) = 2.
(iv) dim(K1,n ) = n 1 (n 2).
We note that there are graphs for which Theorem 2 estimates pd(G1 G2 )
better than Theorem 4 and vice versa. For example Theorem 2 leads to
pd(Kn Pn ) n + 2 while Theorem 4 gives pd(Kn Pn ) n + 1. On
the contrary, if G denotes the unicyclic graph described below, Theorem
2 leads to pd(G G) 12 while Theorem 4 gives pd(G G) 15. In
the above example the unicyclic graph G is composed by fifteen vertices,
where the set 1, 2, 3 form a triangle and the remaining twelve vertices are
leaves: the leaves 4, 5, 6 and 7 are adjacent to 1, the leaves 8, 9, 10 and 11 are
adjacent to 2, and the leaves 12, 13, 14 and 15 are adjacent to 3. In this case
Q
= {{4, 1, 2, 3}, {8}, {12}, {5, 9, 13}, {6, 10, 14}, {7, 11, 15}} is a resolving
partition and S = {4, 5, 6, 8, 9, 10, 12, 13, 14} is a resolving set.
As a direct consequence of above theorem and (1) we deduce the following
interesting result.
Corollary 6. For any connected graphs G1 and G2 ,
pd(G1 G2 ) dim(G1 ) + dim(G2 ) + 1.
One example of graphs for which the equality holds in Corollary 6 (and
also in Corollary 7 (ii)) are the graphs belonging to the family of grid graphs:
pd(Pr Pt ) = 3.
By Remark 5 we obtain the following particular cases of Theorem 4.
Corollary 7. For any connected graph G,
(i) pd(G Kn ) pd(G) + n 1.
(ii) pd(G Pn ) pd(G) + 1.
(iii) pd(G Cn ) pd(G) + 2.
(iv) pd(G K1,n ) pd(G) + n 1.
6

Open problems
1. To prove (or finding a counterexample) that for all pairs of graphs G, H;
dim(G H) dim(G) + dim(H).
2. To provide lower bounds for pd(G H).

Acknowledgments
This work was partly supported by the Spanish Ministry of Science and Innovation through projects TSI2007-65406-C03-01 E-AEGIS and Consolider
Ingenio 2010 CSD2007-0004 ARES.

References
[1] R. C. Brigham, G. Chartrand, R. D. Dutton, P. Zhang, Resolving domination in graphs, Mathematica Bohemica 128 (1) (2003) 2536.
[2] J. Caceres, C. Hernando, M. Mora, I. M. Pelayo, M. L. Puertas, C.
Seara, D. R. Wood, On the metric dimension of Cartesian product of
graphs, SIAM Journal of Discrete Mathematics 21 (2) (2007) 273302.
[3] J. Caceres, C. Hernando, M. Mora, I. M. Pelayo, M. L. Puertas, C.
Seara, On the metric dimension of some families of graphs, Electronic
Notes in Discrete Mathematics 22 (2005) 129133.
[4] G. Chappell, J. Gimbel, C. Hartman, Bounds on the metric and partition dimensions of a graph, manuscript.
[5] G. Chartrand, L. Eroh, M. A. Johnson, O. R. Oellermann, Resolvability in graphs and the metric dimension of a graph, Discrete Applied
Mathematics 105 (2000) 99113.
[6] G. Chartrand, C. Poisson, P. Zhang, Resolvability and the upper dimension of graphs, Computers and Mathematics with Applications 39
(2000) 1928.
[7] G. Chartrand, E. Salehi, P. Zhang, The partition dimension of a graph,
Aequationes Mathematicae (1-2) 59 (2000) 4554.
7

[8] M. Fehr, S. Gosselin, O. R. Oellermann, The partition dimension of

Cayley digraphs Aequationes Mathematicae 71 (2006) 118.
[9] F. Harary, R. A. Melter, On the metric dimension of a graph, Ars Combinatoria 2 (1976) 191195.
[10] T. W. Haynes, M. Henning, J. Howard, Locating and total dominating
sets in trees, Discrete Applied Mathematics 154 (2006) 12931300.
[11] B. L. Hulme, A. W. Shiver, P. J. Slater, A Boolean algebraic analysis
of fire protection, Algebraic and Combinatorial Methods in Operations
Research 95 (1984) 215227.
[12] M. A. Johnson, Structure-activity maps for visualizing the graph variables arising in drug design, Journal of Biopharm. Statist 3 (1993) 203
236.
[13] M. A. Johnson, Browsable structure-activity datasets, Advances in
Molecular Similarity (R. CarboDorca and P. Mezey, eds.) JAI Press
Connecticut (1998) 153170.
[14] S. Khuller, B. Raghavachari, A. Rosenfeld, Landmarks in graphs, Discrete Applied Mathematics 70 (1996) 217229.
[15] R. A. Melter, I. Tomescu, Metric bases in digital geometry, Computer
Vision Graphics and Image Processing 25 (1984) 113121.
[16] V. Saenpholphat, P. Zhang, Conditional resolvability in graphs: a survey, International Journal of Mathematics and Mathematical Sciences
38 (2004) 19972017.
[17] P. J. Slater, Leaves of trees, Proc. 6th Southeastern Conference on Combinatorics, Graph Theory, and Computing, Congressus Numerantium 14
(1975) 549559.
[18] I. Tomescu, Discrepancies between metric and partition dimension of a
connected graph, Discrete Mathematics 308 (2008) 50265031.