AKCE International Journal of Graphs and Combinatorics

ISSN: 0972-8600 (Print) 2543-3474 (Online) Journal homepage: www.tandfonline.com/journals/uakc20

Algorithmic complexity of secure connected

domination in graphs

J. Pavan Kumar, P. Venkata Subba Reddy & S. Arumugam

To cite this article: J. Pavan Kumar, P. Venkata Subba Reddy & S. Arumugam (2020) Algorithmic
complexity of secure connected domination in graphs, AKCE International Journal of Graphs
and Combinatorics, 17:3, 1010-1013, DOI: 10.1016/j.akcej.2019.08.012

To link to this article: https://doi.org/10.1016/j.akcej.2019.08.012

© 2020 The Author(s). Published with

license by Taylor & Francis Group, LLC

Published online: 28 May 2020.

Submit your article to this journal

Article views: 844

View related articles

View Crossmark data

Citing articles: 4 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=uakc20
AKCE INTERNATIONAL JOURNAL OF GRAPHS AND COMBINATORICS
2020, VOL. 17, NO. 3, 1010–1013
https://doi.org/10.1016/j.akcej.2019.08.012

Algorithmic complexity of secure connected domination in graphs

J. Pavan Kumara , P. Venkata Subba Reddya , and S. Arumugamb
a
Department of Computer Science and Engineering National Institute of Technology Warangal, Warangal, Telangana, India; bn-CARDMATH,
Kalasalingam Academy of Research and Education, Krishnankoil, Tamilnadu, India

ABSTRACT KEYWORDS
Let G ¼ ðV, EÞ be a simple, undirected, and connected graph. A connected (total) dominating set Domination; secure
S  V is a secure connected (total) dominating set of G, if for each u 2 V n S, there exists v 2 S domination; secure
such that uv 2 E and ðS n fvgÞ [ fug is a connected (total) dominating set of G. The minimum car- connected domination;
w[2]-hard
dinality of a secure connected (total) dominating set of G denoted by csc ðGÞðcst ðGÞÞ, is called the
secure connected (total) domination number of G. In this paper, we show that the decision prob-
lems corresponding to secure connected domination number and secure total domination number
are NP-complete even when restricted to split graphs or bipartite graphs. The NP-complete reduc-
tions also show that these problems are w[2]-hard. We also prove that the secure connected dom-
ination problem is linear time solvable in block graphs and threshold graphs.

1. Introduction fug is a TDS in G. The minimum cardinality of a STDS in

G is called the secure total domination number of G and is
Let G(V, E) be a simple, undirected, and connected graph.
denoted by cst ðGÞ: We need the following theorems.
For graph theoretic terminology we refer to [8]. For a vertex
v 2 V, the open neighborhood of v in G is NG ðvÞ¼ Theorem 1. ([2]) Let G be a connected graph of order n.
fu 2 V : uv 2 Eg, the closed neighborhood of v is defined as Then csc ðGÞ ¼ 1 if and only if G ¼ Kn :
NG ½v ¼ NG ðvÞ [ fvg: If S  V, then the open neighborhood Theorem 2. ([2]) Let G be a connected graph of order n  3.
of S is the set NG ðSÞ ¼ [v2S NG ðvÞ: The closed neighborhood Let L(G) and S(G) be the set of pendant and support vertices
of S is NG ½S ¼ S [ NG ðSÞ: Let S  V: Then a vertex w 2 V of G respectively. Let X be a secure connected dominating set
is called a private neighbor of v with respect to S if N½w \ of G. Then (i) LðGÞ  X and SðGÞ  X
S ¼ fvg: If further w 2 V n S, then w is called an external
private neighbor (epn) of v. (ii) No vertex in LðGÞ [ SðGÞ is an X-defender.
A subset S of V is a dominating set (DS) in G if for every
u 2 V n S, there exists v 2 S such that uv 2 E: The domin- Proposition 1. ([3]) Let S be a CDS in G. Then S is a SCDS
ation number of G is the minimum cardinality of a DS in G in G if and only if the following conditions are satisfied.
and is denoted by cðGÞ: A set S  V is said to be a secure
dominating set (SDS) in G if for every u 2 V n S there exists (i) epnðv, SÞ ¼ ; for all v 2 S:
v 2 S such that uv 2 E and ðS n fvgÞ [ fug is a dominating (ii) For every u 2 V n S, there exists v 2 S \ NG ðuÞ such that
set of G. We say that v S-defends u or u is defended by v VðCÞ \ NG ðuÞ 6¼ ; for every component C of G½S n fvg:
[4]. The minimum cardinality of a SDS in G is called the
secure domination number of G and is denoted by cs ðGÞ: A Proposition 2. ([2]) Let G be a non-complete connected
dominating set S is said to be a connected dominating set graph and let S be a secure connected dominating set in G.
(CDS), if the induced subgraph G½S is connected. A CDS S Then the set S n fvg is a dominating set for every v 2 S. In
is said to be a secure connected dominating set (SCDS) in G particular, 1 þ cðGÞ  csc ðGÞ:
if for each u 2 V n S, there exists v 2 S such that uv 2 E
and ðS n fvgÞ [ fug is a CDS in G. The minimum cardinal- 2. Main results
ity of a SCDS in G is called the secure connected domination We first determine the value of csc ðGÞ for two families of
number of G and is denoted by csc ðGÞ: A dominating set S graphs.
is said to be a total dominating set (TDS), if the induced
subgraph G½S has no isolated vertices. A TDS S is said to be Theorem 3. Let Wn ¼ v1 þ Cn be the wheel of order n þ 1
a secure total dominating set (STDS) of G, if for each u 2 where n  3. Let G be the graph obtained from Wnþ1 by sub-
V n S, there exists v 2 S such that uv 2 E and ðS n fvgÞ [ dividing all the edges of Cn. Then csc ðGÞ ¼ n þ 1:

CONTACT J. Pavan Kumar [email protected] Department of Computer Science and Engineering National Institute of Technology Warangal, Warangal,
Telangana, India.
ß 2020 The Author(s). Published with license by Taylor & Francis Group, LLC
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly cited.
 AKCE INTERNATIONAL JOURNAL OF GRAPHS AND COMBINATORICS 1011

Proof. Let VðGÞ ¼ fv1 , v2 , :::, v2nþ1 g, dðv2nþ1 Þ ¼ n, dðvi Þ ¼

We now proceed to present results on algorithmic aspects
2 if i is even such as NP-comppleteness and linear time algorithm for
and
3 otherwise some classes of graphs.
Nðvi Þ ¼ fvi1 , viþ1 g if i is even. Then S ¼ fvi : i is oddg Secure Connected Domination Problem (SCDM)
is a SCDS of G. Hence csc ðGÞ  n þ 1: Instance: A connected graph G and a positive integer l:
Now let D be any csc -set of G. If v2nþ1 62 D or if v2nþ1 2 Question: Does there exist a SCDS of size at
D and defends a vertex vi , then we get a connected domi- most l in G?
nating set D1 ofG such that jD1 j ¼ jDj and v2nþ1 62 D1 : The proof is by reduction from the Domination problem
Hence jDj ¼ jD1 j  2n  2, which is a contradiction. Thus (DM), which is NP-complete [5].
v2nþ1 2 D and v2nþ1 does not defend any other vertex. Now Domination Problem (DM)
let vi 2 D for some iwherei is even. Since G½D is connected, Instance: A graph G and a positive integer k.
one of vi1 or viþ1 is in D: Also if vi 62 D for all even Question: Does there exist a DS of size at most k in G?
i, then vi 2 D for all odd i: Hence csc ðGÞ ¼ jDj  n þ 1: w
Theorem 6. SCDM is NP-complete.
Theorem 4. For the book graph Bn ¼ K1, n w K2 , we have Proof. It can be easily verified that SCDM is in NP. Now let
csc ðBn Þ ¼ n þ 2: G ¼ ðV, EÞ be a graph and let k be a positive integer. Let G
be the graph with VðG Þ ¼ V [ fxg and EðG Þ ¼ E [ fðu, xÞ
Proof. Let S1 and S2 be the two copies of K1, n in Bn : Let VðS1 Þ ¼ : u 2 Vg and let l ¼ k þ 1: Clearly, G can be constructed
fv1 ,v2 ,:::,vnþ1 g and VðS2 Þ ¼ fw1 ,w2 ,:::,wnþ1 g: Let v1 and w1 be from G in polynomial time.
the central vertices of S1 ,S2 , respectively. Let vi wi 2 EðBn Þ: Clearly Now if D is a dominating set of G with jDj  k, then
VðS1 Þ [ fw1 g is an SCDS of Bn : Hence csc ðBn Þ  n þ 2: S ¼ D [ fxg is an SCDS of G : Conversely, let S be an
Now let D be any csc -set of Bn : Since D is connected, SCDS of G with jS j  k þ 1: If x 2 S, then it follows from
either v1 or w1 is in D: If w1 2 D and v1 62 D, then fw2 , w3 , Proposition 1 that epnðx, SÞ ¼ ;: Therefore, every vertex u 2
:::, wnþ1 , v2 , v3 , :::, vnþ1 g  D: Thus, jDj  2n þ 1 which is a VðG Þ n S is adjacent to a vertex in S n fxg: Hence S n fxg is
contradiction. Hence v1 , w1 2 D: Now if both wi and vi are a DS of size at most k in G: If x 62 S, Proposition 2, the set
not in D for some i  2, then G½ðD n fw1 gÞ [ S n fvg, for any v 2 S, is a DS of size at most k in G: w
fwi g and GðD n fv1 gÞ [ fvi g are disconnected. Hence jD \
fwi , vi gj  1 for any i  2 and csc ðBn Þ ¼ jDj  n þ 2: Thus, Next, we define the decision version of total domination
csc ðBn Þ ¼ n þ 2: w and secure total domination problems as follows.
Total Domination Problem (TDM)
Instance: A simple, undirected graph G without isolated
Theorem 5. Let G ¼ Pn wP2 where n  3: Then csc ðGÞ ¼
vertices and a positive integer r.
n þdn3e:
Question: Does there exist a TDS of size at most r in G?
Secure Total Domination Problem (STDM)
Proof. Let P ¼ ðv1 , v2 , :::, vn Þ and Q ¼ ðw1 , w2 , :::, wn Þ be two
Instance: A simple, undirected and connected graph G
copies of Pn in G such that vi wi 2 EðGÞ: Let V1 ¼ fv1 , v2 ,
:::, vn g and V2 ¼ fw1 , w2 , :::, wn g: Then S ¼ V1 [ fwi : i  2 and a positive integer m:
ðmod3Þg is a SCDS of G: Hence csc ðGÞ  n þ dn3e: Question: Does there exist a STDS of size at most minG?
Let D be any csc -set of G: If vi , wi 62 D for some Theorem 7. STDM is NP-complete.
i, where 2  i  n  1, then G½D is disconnected, which is a
contradiction. Hence at least one of vi , wi is in D, where 2  Proof. It is clear that STDM is in NP. The reduction given
i  n  1: If both v1 and w1 are not in D, then G½ðD n fv2 gÞ [ in the proof of Theorem 6 shows that STDM is NP-com-
fv1 g and G½ðD n fw2 gÞ [ fw1 g are disconnected, which is a plete. w

contradiction. Hence v1 or w1 is in D: Similarly, wn or vn is in We now give NP-completeness results even when

D: We now claim that D \ V1 is a dominating set of P: restricted to bipartite graphs or split graphs. We formulate
Suppose there exists a vertex vi such that vi is not dominated the SCDM for bipartite graphs as follows.
by D \ V1 : Then wi 2 D and G½ðD n fwi gÞ [ fvi g is discon- Secure Connected Domination Problem for Bipartite
nected, which is a contradiction. Hence D \ V1 is a dominat- Graphs (SCDB)
ing set of P: Similarly D \ V2 is a dominating set of Q: Now Instance: A connected bipartite graph G ¼ ðV1 , V2 , EÞ
suppose D \ V1 ˆV1 and D \ V2 ˆV2 : If three consecutive and a positive integer r.
vertices of P say, vi , viþ1 , viþ2 are not in D, then Question: Does there exist a SCDS of size at most r in G?
wi , wiþ1 , wiþ2 2 D: However, G½ðD n fwiþ1 gÞ [ fviþ1 g is dis-
connected, which is a contradiction. Now suppose vi , viþ1 62 Theorem 8. SCDB is NP-complete.
D: Then vi1 , viþ2 , wi , wiþ1 2 D: Now since G½D is connected, Proof. It can be seen that SCDB is in NP. We transform an
it follows that wi1 , wiþ2 2 D: HenceðD n fwi , wiþ1 gÞ [ instance of SCDM problem to an instance of SCDB as fol-
fvi , viþ1 g is also a SCDS of G: Thus by repeating the above lows. Given a graph G, we construct a graph
process we get a SCDS of G, D1 such that jD1 j ¼ jDj, D1 \ G ðV1 , V2 , EÞ where V1 ðG Þ ¼ V [ fp, qg, V2 ðG Þ ¼ V 0 ðGÞ [
V1 ¼ V1 and D1 \ V2 is a dominating set of Q: Thus, jDj ¼ fx, yg, here V 0 ðGÞ is another copy of V such that if u and v
jD1 j  n þ dn3e: Therefore, csc ðGÞ ¼ n þ dn3e: w
are two vertices in V then the corresponding vertices in
1012 J. P. KUMAR ET AL.

a a

b b

c c
a d e

d d

b c
e e
(a) Graph G

p x

q y

(b) Graph G∗
Figure 1. Construction of G from G:


V 0 ðGÞ are labeled as u0 and v0 , and EðG Þ consists of (i) Q and an independent set I, we construct a split graph G
edges uv0 and u0 v for each edge uv 2 E; (ii) edges of the with a clique Q and an independent set I  as follows:
form uu0 for each vertex u 2 V; and (iii) edges of the form
VðG Þ ¼ V [ fx, yg, and
ux and uy for every vertex u 2 V1 ðG Þ: Clearly G is a
bipartite graph and can be constructed from G in polyno- EðG Þ ¼ E [ fxu : u 2 Vg [ fxyg:
mial time (Figure 1). Note that G is a split graph, where Q ¼ Q [ fxg and I  ¼
Next, we show that G has a SCDS of size at most r if and I [ fyg and G can be constructed from G in polyno-
only if G has a SCDS of size at most r þ 2: If S is a SCDS mial time.
of G withjSj  r, then it can be easily verified that S ¼ Now let S be a DS of G with jSj  k: Then S ¼ S [ fx, yg is
S [ fx, yg is a SCDS of G with jS j  r þ 2: a SCDS of G with jS j  k þ 2: Conversely, let S be a SCDS
Conversely, let S be an SCDS of G and jS j  r þ 2: Since of G with jS j  k þ 2: It follows from Proposition 2 that
xandy are the only vertices in S which defend pandq, it fol- x, y 2 S : Clearly S0 ¼ S n fx, yg is a DS of G with jS0 j  k: w
lows that at least one of them must be in S : If x 2 S and y 62
S , then G ½ðS n fxgÞ [ fpg is disconnected, which is a
Theorem 11. STDM is NP-complete for split graphs.
contradiction. Hence x, y 2 S : Let S0 ¼ S n
fx, y, p, qg and S ¼ ðS [ fv : v 2 S \ V ðGÞgÞ n
00 0 0 0 0
Proof. It is clear that STDM for split graphs is in NP. The
fv0 : v0 2 S0 \ V 0 ðGÞg: Clearly S00 forms a SCDS of size at most reduction given in the proof of Theorem 10 shows that
r in G: w
STDM is NP-complete for split graphs. w

Theorem 9. STDM is NP-complete for bipartite graphs. Since the Domination problem is w[2]-complete for split
graphs [7] and the reductions in Theorems 10 and 11 are in
Proof. It is clear that STDM for bipartite graphs is in NP. the function of the parameter c, the following two corolla-
The reduction given in the proof of Theorems 8 shows that ries are immediate.
STDM is NP-complete for bipartite graphs. w
Corollary 3. SCDM is w[2]-hard in split graphs.
Since the Domination problem is w[2]-complete for
bipartite graphs [7] and the reductions in Theorems 8 and 9 Corollary 4. STDM is w[2]-hard in split graphs.
are in the function of the parameter l, the following two In the next two theorems we prove that csc ðGÞ can be
corollaries are immediate. computed in linear time for block graphs and thresh-
old graphs.
Corollary 1. SCDM is w[2]-hard in bipartite graphs.
Let G ¼ ðV, EÞ be a connected graph. A vertex v is called
Corollary 2. STDM is w[2]-hard in bipartite graphs. a cut-vertex of G if G  v is a disconnected graph. A graph
It has been shown that the DM and the TDM as NP- G with no cut-vertex is called a block. A block B of a graph
is a maximal connected induced subgraph of G such that B
complete even when restricted to split graphs [1].
has no cut-vertex. In block B, vertices which are not cut ver-
Theorem 10. SCDM is NP-complete for split graphs. tices of G are called block vertices. A graph G is called a
block graph if all its blocks are cliques.
Proof. It is known that SCDM is a member of NP. We
reduce DM for split graphs to SCDM for split graphs. Given Definition 1. A graph G is called a block graph if all the
a split graph G whose vertex set is partitioned into a clique blocks of G are cliques.
 AKCE INTERNATIONAL JOURNAL OF GRAPHS AND COMBINATORICS 1013

Theorem 12. Let G be a block graph having r blocks and k is a support vertex. Hence no vertex of C n fug is D-defended,
cut vertices. Then csc ðGÞ ¼ k þ r  r0 , where r0 is the number which is a contradiction. Hence csc ðGÞ ¼ jDj  2 þ l: Thus
of blocks such that all vertices of the block are cut vertices. csc ðGÞ ¼ 2 þ l: w

Proof. Let A denote the set of all cut vertices of G. Let Theorem 14. SCDM is linear time solvable for thresh-
B1 , B2 , :::, Br0 , Br0 þ1 , :::, Br be the blocks of G where every ver- old graphs.
tex of Bi is a cut vertex of G if 1  i  r0 : Let T ¼
fvi : 1  i  r  r0 and vi is a non-cut vertex of Br0 þi g: Let Proof. Since the ordering of the vertices of the clique in a
S ¼ A [ T:SinceA contains all cut-vertices of G, it follows threshold graph can be determined in linear time [6], the
that G½S is connected. Also if v 2 V n S, then v is not a result follows. w
cut-vertex. Now there exists a vertex u 2 T such that uv 2
EandðS n fvgÞ [ fug is a CDS of G: Hence, S is a SCDS of
G: Therefore, csc ðGÞ  k þ r  r0 : 3. Conclusion
Now let D be any csc -set of G: Since G½D is connected, D
A: Further, a cut-vertex cannot defend any other vertex In this paper, it is shown that secure connected (total) dom-
and hence D contains at least one non-cut vertex from each ination problem is NP-complete even when restricted to
block Bi where r0 þ 1  i  r: Hence csc ðGÞ ¼ jDj  jSj ¼ k bipartite graphs, or split graphs. Since split graphs form a
þr r0 : Thus csc ðGÞ ¼ k þ r  r0 : w proper subclass of chordal graphs, these problems are also
NP-complete for chordal graphs. We have proved that
Corollary 5. Let G be a block graph with r blocks and exactly secure connected domination problem is linear time solvable
one cut-vertex. Then csc ðGÞ ¼ r þ 1: for block graphs and threshold graphs. It will be interesting
to investigate the algorithmic complexity of secure con-
Corollary 6. For any tree T with n vertices, csc ðTÞ ¼ n: nected (total) domination problem for subclasses of chordal
and bipartite graphs.
Proof. Here r ¼ n  1, r0 ¼ n  1  l and k ¼ n  l where l
is the number of leaves in T: Therefore, csc ðTÞ ¼ n: w
Disclosure statement
Corollary 7. SCDM is linear time solvable for block graphs. No potential conflict of interest was reported by the authors.

Proof. Since number of blocks and number of cut-vertices of

block graph can be determined in linear time, the result fol- ORCID
lows. w
J. Pavan Kumar http://orcid.org/0000-0001-6936-2358
P. Venkata Subba Reddy http://orcid.org/0000-0002-0972-1141
Definition 2. A graph G ¼ ðV, EÞ is called a threshold S. Arumugam http://orcid.org/0000-0002-4477-9453
graph if there is a real number t and a real number wðvÞ for
every vP2 V such that a set S  V is independent if and
References
only if v2S wðSÞ  t:
[1] Bertossi, A. A. (1984). Dominating sets for split and bipartite
Threshold graphs considered are assumed to be non- graphs. Inf. Process. Lett. 19(1): 37–40.
complete and connected. We use the following characteriza- [2] Cabaro, A. G., Canoy, S. S., Jr., Aniversario, I. S. (2014). Secure
tion of threshold graphs given in [6] to prove that secure connected domination in a graph. Int. J. Math. Anal. 8:
connected domination number can be computed in linear 2065–2074.
time for threshold graphs. [3] Cabaro, A. G, Canoy, S. R., Jr. (2015). Secure connected domi-
A graph G is a threshold graph if and only if it is a split nating sets in the join and composition of graphs. Int. J. Math.
graph and for split partition ðC, IÞ of V, there is an order- Anal. 9(25): 1241–1248.
[4] Cockayne, E. J., Grobler, P. J. P., Grundlingh, W. R.,
ing ðx1 , x2 , :::, xp Þ of vertices of C such that N½x1   N½x2   Munganga, J, van Vuuren, J. H. (2005). Protection of a graph.
:::  N½xp , and there is an ordering ðy1 , y2 , :::, yq Þ of the Utilitas Math. 67: 19–32.
vertices of I such that Nðy1 Þ Nðy2 Þ :::Nðyq Þ: [5] Garey, M. R, Johnson, D. S. (1979). Computers and
Intractability: A Guide to the Theory of NP-Completeness. New
Theorem 13. Let G be a connected threshold graph. Then
York, NY: Freeman.
csc ðGÞ ¼ 2 þ l, where l is the number of pendant vertices. [6] Mahadev, N, Peled, U. (1995). Threshold Graphs and Related
Proof. Let S ¼ fxp , xp1 g [ fv 2 I : v 2 Nðxp Þ n Nðxp1 Þg: Topics. In: Annals of Discrete Mathematics, Vol. 56. North
Holland: The Netherlands.
Clearly G½S is a star with center xp : Also every vertex v 2
[7] Raman, V, Saurabh, S. (2008). Short cycles make W-hard prob-
V n S is defended by xp1 and G½ðS n fxp1 gÞ [ fvg is con- lems hard: FPT algorithms for W-hard problems in graphs with
nected. Thus, S is a SCDS of G:Hencecsc ðGÞ  2 þ l: no short cycles. Algorithmica 52(2): 203–225.
Let D be any csc -set of G. It follows from Theorem 2 that [8] West, D. B. (2001). An Introduction to Graph Theory, Vol. 2.
jDj  l þ 1:If jDj ¼ l þ 1, then exactly one vertex of C say, u Upper Saddle River, NJ: Prentice Hall.