1. Introduction
In general, we use the standard terminology and notation of graph theory; see [
1]. Let
be an undirected, simple graph with vertex set
and edge set
. Two vertices
of
G are adjacent, or neighbors, if
is an edge in
G. The degree of a vertex
is denoted by
and the maximum degree of
G is denoted by
. The set of all vertices of maximum degree is denoted by
. The number
is the minimum degree of
G. A vertex of degree 1 is a leaf and
. The set
is an open neighborhood of a vertex
v and the set
is a closed neighborhood. The distance between vertices
in
G is denoted by
. Let
, then for
the distance is defined as
. A maximal, with respect to a set inclusion, connected subgraph of
G is called a component of
G. A cutvertex in a graph G is a vertex whose removal increases the number of connected components of
G.
Graph theory has been an essential area of mathematical research due to its wide range of applications and theoretical richness. This paper explores the concept of domination in graphs, a topic that has gained significant attention since its introduction in the 19th century by de Jaenish [
2] in chessboard problems and formalized by Berge [
3] and Ore [
4] in the 1960s. A dominating set (DS) of a graph
G is a subset
such that every vertex
is adjacent to at least one vertex
. In particular,
is also dominating.
Over the decades, various generalizations and modifications of dominating sets have been proposed, including the concept of secondary domination, introduced by Hedetniemi et al. in 2008 [
5]. Let
. A subset
is called a
-dominating set (or
-DS) if each vertex
has a neighbor in
D and another vertex in
D at a distance of at most
k from
v. In particular,
is also a
-DS for every
. In [
5], it was proved that the study of a
-DS is meaningful only for
.
Among the special cases,
-dominating sets, also known as 2-dominating or double dominating sets, have been extensively studied; see [
6,
7,
8,
9]. However, the study of
-dominating sets has appeared as a natural extension, with interesting properties and computational challenges. The concept of
-dominating sets has inspired significant research in recent years. Michalski et al. [
10], Kayathri and Vallirani [
11], and Raczek [
12,
13] have explored its structural properties and computational complexity. Additionally, independent dominating sets in graphs and their products were studied in [
5,
14,
15]. Furthermore, the notion of proper
-dominating sets was introduced by Michalski in [
16], which improved the study of these subsets by excluding cases that are
-dominating sets. A subset
is a proper
-DS if
D is a
-DS and not a
-DS, i.e.,
D is
-dominating, and there exists a vertex
such that
x has exactly one neighbor in
D, and there is a vertex
such that
. Since the set
is a
-DS, it is not the proper
-DS. The properties of proper
-DS and the problem of its existence were considered in the papers [
10,
17,
18].
Motivated by the above, Bednarz and Pirga studied
-dominating sets and introduced the concept of a perfect
-dominating set, a special case of a proper
-DS (see [
19]). A subset
is a perfect
-dominating set of
G (or
-PDS) if every vertex
has exactly one neighbor in
D, and there exists a vertex
such that
. In this case, we say that
x is perfect
-dominated by
D (abbreviated as
x is
-PD).
A proper -DS only requires the existence of at least one vertex outside D that satisfies the domination condition, meaning that there exists a vertex such that . A perfect -DS imposes this condition on all vertices outside D. Thus, every perfect -dominating set is a proper -dominating set, but not every proper -dominating set is perfect, as it may fail to enforce this condition for all vertices outside D.
It is evident that if
D is a perfect
-dominating set, then it is also a proper
-DS, and
cannot be a perfect
-dominating set. In [
19], the problem of the existence of
-PDS in graphs was considered. Specifically, the focus was on the class of graphs with at most two vertices of maximum degree, analyzing the structural conditions under which such sets exist. First, the case where the maximum degree of a vertex was
was considered.
Theorem 1 ([
19])
. Let G be a connected n-vertex graph of the maximum degree , . If G has a -PDS, then G has the unique vertex of maximum degree . A complete characterization of graphs with which admit a -PDS was also presented.
Theorem 2 ([
19])
. Let G be a connected n-vertex graph, with the unique vertex of maximum degree . G has a -PDS if and only if the vertex of maximum degree is a cutvertex. Next, the existence of a -PDS in graphs where the maximum degree of a vertex was was examined. Complete solutions, along with necessary and sufficient conditions, were provided for graphs with one or two such vertices.
This paper is a continuation of the research presented in [
19]. Necessary and sufficient conditions for the existence of
-PDS will be proven in the case where there are three vertices with a maximum degree of
. We will consider cases where these vertices are adjacent or nonadjacent and whether their neighborhoods are identical or distinct.
It is worth emphasizing that dominating sets have a wide range of applications. In [
13], Raczek described numerous applications of dominating sets, including
-dominating sets. Dominating sets in networks represent a subset of special nodes (e.g., servers or service providers) that ensure direct resource availability. The domination number is the smallest possible size of such a subset. Applications of domination parameters include environmental sensors, vehicular communication, road safety, healthcare, peer-to-peer messaging, disaster rescue, and defense systems. Another application is the relationship between the minimum dominating set problem and other NP-hard problems, such as the vertex k-center problem, which plays a crucial role in optimization and facility location theory (see [
20]).
In some networks, higher reliability is required, where a backup node can take over if a primary node fails. In such cases, -domination can be useful, as backup nodes can be located at a distance of 1 or 2. While backup nodes at a distance of 2 may operate with reduced performance, they ensure that critical tasks are covered in emergencies, similar to how emergency numbers function during poor signal coverage.
An example application of -dominating sets is in pandemic scenarios, such as ensuring optimal placement of medical care facilities. Towns are represented as vertices, and edges connect towns within a specified distance. The minimum -dominating set identifies the best locations for medical facilities, ensuring every town has access to a primary facility within a certain distance and a backup facility within twice that distance.
A similar application can be identified in the case of perfect -domination. Another example of such an application is related to wars or natural disasters. In such situations, perfect -dominating sets can ensure that each point (e.g., a town or a strategic location) has access to a primary shelter or resource hub, with an additional backup shelter located at a greater distance. This ensures that in emergencies, all points have a safe and reliable refuge, even if one shelter becomes inaccessible due to unforeseen circumstances.
2. Perfect -Dominating Sets in Graphs with Maximum Degree
We will consider cases where . Let and . In this paper, we will consider the following cases:
- (i)
x and y are not adjacent, and z is adjacent to both x and y, and , , ;
- (ii)
and z are adjacent, and ;
- (iii)
and z are adjacent, and , , ;
- (iv)
and z are adjacent, and , , .
It was proved in [
19] that if graph
G satisfies
, then
G has a
-PDS. Therefore, in our subsequent considerations, we assume that
.
Theorem 3. Let G be an n-vertex connected graph, and , , such that and are not adjacent and . A vertex z is adjacent to x and to y and and . A graph G has a -PDS if and only if there exists a unique vertex of degree 2 adjacent to vertices x and y.
Proof. Let us assume that the assumptions of Theorem 3 are satisfied. Let R be a set defined as . Then, there is no vertex in R that is adjacent to all other vertices in this set. Otherwise, if at least one vertex in R is adjacent to all other vertices in this set, then it has a degree , which contradicts the assumption that .
If
, then no graph
G satisfies the above assumptions. Let
and
,
, such that
and
are not adjacent, and
. Suppose that there exists a unique vertex
such that
and it is adjacent to
x and
y. Because
and a vertex
is adjacent to
x and
y, it implies that
u is the only vertex of a graph
G that is not adjacent to a vertex
z; otherwise,
. Hence,
and
. Then, sets
or
are
-PDS of a graph
G. An example of such a graph for
, where
and
, is shown in
Figure 1.
Now, suppose that a graph G has a -PDS and let us denote it by D. It is clear that and also only one of the vertices from the set is in D; otherwise, D is a -dominating set. Assume that . Hence all vertices from a set R and a vertex z are dominated by x. Moreover, , otherwise vertices are -dominated by D, which is a contradiction. From the fact that and and and , it follows that there exists a unique vertex u in a graph G that is not adjacent to z; otherwise, . A vertex u is adjacent to x and to y; otherwise, . Hence, and . Therefore, . Because u is adjacent to x and to y, it follows that . If , then there exists a vertex adjacent to u. Therefore, v is -dominated by D, a contradiction. Thus, the theorem is proved. ☐
Theorem 4. Let G be an n-vertex connected graph, and , , such that and are adjacent, . We denote a set of vertices by . Let u be a unique vertex that is not adjacent to vertices . A graph G has a -PDS if and only if there is a vertex such that every is adjacent to either u or v, and , and every vertex from is not adjacent to at least one vertex from and
- (i)
and ;
- (ii)
and every vertex from is adjacent to a vertex from .
Proof. Let us assume that the assumptions of Theorem 4 are satisfied. Let
G be an
n-vertex connected graph,
,
,
, such that
. If
, then there is no graph satisfying the above assumptions. Let
u be a unique vertex that is neither adjacent to
x,
y, nor
z. Let
. Then, there is no vertex adjacent to all other vertices in
B; otherwise, there is a vertex
such that
, which contradicts the cardinality of a set
. Since
, it follows that a vertex
is adjacent to all vertices from set
B. Assume that there is a vertex
such that every
is adjacent to either
u or
v, and
, and every vertex from
is not adjacent to at least one vertex from
, and conditions (i) or (ii) are valid. We will prove that a set
is a
-PDS of a graph
G. Vertices
are dominated by
. Because
u is neither adjacent to
x,
y, nor
z, it follows that
. Hence
are
-PD by
D. If a vertex
, then
is dominated by
u. By the connectivity of
G, there is a path
; hence,
is
-PD by
D. If
, then
is dominated by
v. If a condition (i) is valid, then there is a path
in graph
G and
is
-PD by
D. Since
, then
. If condition (ii) is valid, then there exists a vertex
such that
. Hence, there exists a path
in graph
G, and
is
-PD by
D. From the fact that
, all vertices in
B are not
-dominated by
D. Because there is no vertex adjacent to all other vertices in
B, and every vertex in
is not adjacent to at least one vertex in
, it follows that
for all
. An example graph for
, where
, and
for case (i) is presented in
Figure 2.
Conversely, assume that G has a -PDS, which we denote by D. Firstly, we prove some claims.
(1) , and , and .
If , then vertices from set B are -dominated, a contradiction. Analogously we can prove that and .
(2)
It is obvious because a vertex u is the only vertex in a graph G such that ; otherwise, are not -PD by D.
(3) , , and .
Assume that . Then, there exists a vertex that is adjacent to a vertex u and to x. Hence, is -dominated by D because by (2). Analogously, we can prove that and .
(4) .
If , then , or , which contradicts (3). Assume that . Vertices are adjacent to every vertex in set B; hence, are -dominated by D, a contradiction.
From claims (1)–(4), it follows that there exists a vertex such that . Hence, . Vertices are -PD by D. Since , it implies that every vertex is adjacent to either u or v, and ; otherwise, there exists a vertex such that either w is not dominated or w is -dominated by D. Moreover, every vertex is not adjacent to at least one vertex in ; otherwise, , which contradicts . Since , a vertex u is adjacent to at least two vertices in set B. If u is adjacent to a vertex , then a vertex u must be adjacent to two other vertices in set B; otherwise, , which contradicts . Hence, , and condition (i) is valid. If u is not adjacent to v, then every vertex in is adjacent to a vertex in ; otherwise, there exists a vertex that is dominated by v but is not -PD by D. Hence, condition (ii) is valid. Thus, the theorem is proved. ☐
Theorem 5. Let G be an n-vertex connected graph, , and , , such that , where are adjacent and and , . Let us denote by H a set of vertices of graph G such that . Let u be a unique vertex not adjacent to x or y but adjacent to z. A graph G has a -PDS if and only if there exists a vertex such that every is adjacent to u or v, and , and every vertex in is not adjacent to at least one vertex in .
Proof. Let G be an n-vertex connected graph. Let , where are adjacent and and , . If , then no graph satisfies the assumptions of this theorem. Let and u be a unique vertex adjacent to a vertex z. Let . Then, there is no vertex adjacent to all other vertices in H. Otherwise, there exists a vertex in H with degree , which contradicts the cardinality of the set .
Suppose there exists a vertex
such that every vertex
is adjacent to
u or
v, and
, and every vertex in
is not adjacent to at least one vertex in
. We will prove that a set
is a
-PDS of a graph
G. Vertices
x and
y are dominated by
and
. Hence,
are
-PD by
D. Since
and
and
, a vertex
z is not adjacent to exactly one vertex in a set
H. If
, then
z is not adjacent to a vertex
v. Hence,
z is dominated by
u, and there exists a path
, then
z is
-PD by
D. If
, then
z is not adjacent to one vertex
. Every vertex
is dominated by
u or
v and is not
-dominated because
. If
, then there exists a path
; hence,
is
-PD by
D. If
, then there exists a path
; hence,
is
-PD by
D. Since every vertex in
is not adjacent to at least one vertex in
, its degree is less than or equal to
. An example graph for
, where
, and
, and
is presented in
Figure 3.
Conversely, assume that a set D is -PDS of a graph G. We will prove the following claims.
(1) .
If , then vertices in a set H are -dominated, a contradiction.
(2) .
Because and , a vertex u is the only vertex in G not adjacent to vertices x and y. Hence, ; otherwise, are not -PD by D.
(3) and .
This can be proved analogously to case (3) in Theorem 4.
(4) .
If , then x or y is in D, contradicting (3). If , then vertices x and y are -dominated by D, because they are adjacent to every vertex in a set H, a contradiction.
From the above cases, it follows that there exists a vertex such that . Hence, . It is obvious that every vertex is adjacent to either u or v, and ; otherwise, would be either - dominated by D or not dominated at all. Since H contains no vertex adjacent to all other vertices in H, every vertex satisfies the condition . Furthermore, because and a vertex z is adjacent to a vertex u, there must exist at least one vertex such that is adjacent to u. Consequently, every vertex is not adjacent to at least one vertex in a set . Otherwise, there would exist a vertex such that , which would contradict the assumption that . Vertices x and y are -PD by D because x and y are adjacent to a vertex v. By (2), it follows that . Thus, the theorem is proved. ☐
Theorem 6. Let G be an n-vertex connected graph with , and , such that , where are adjacent. If , , and , then there is no -PDS in G.
Proof. Assume that G is an n-vertex connected graph with , and , such that , where are adjacent. Since , , and , it follows that there exist vertices such that is adjacent to both x and z but not to y, and is adjacent to both y and z but not to x, and u is adjacent to both x and y but not to z.
By contradiction, assume that G has a -PDS, denoted by D. Hence, every vertex must have exactly one neighbor in D and another vertex in D at a distance of exactly 2. If , then all vertices in G are dominated by x, except for , which is not adjacent to x. The vertex cannot belong to D because, in that case, y would be -dominated. Hence, is not in the set D and is not dominated, leading to a contradiction. In the same way, we can prove that . If , then the vertices x and z are dominated. The only vertex at a distance of 2 from x is , which cannot be in D because, in that case, vertex z would be -dominated. Thus, there is no vertex that belongs to D without -dominating other vertices. Therefore, .
Finally, consider any vertex which belongs to D. Then, there is no vertex in D at a distance of 2 from or z such that the remaining vertices are not -dominated.
Taking all the considered cases into account, we conclude that a -PDS cannot exist in G. ☐