Domination Parameters in Hypertrees and Sibling trees
Indra Rajasingh, R. Jayagopal, R. Sundara Rajan

TL;DR
This paper investigates various domination parameters, including locating and total domination numbers, specifically for hypertrees and sibling trees, providing exact values and insights into their structural properties.
Contribution
It determines the exact values of domination, total domination, locating-domination, and locating-total domination numbers for hypertrees and sibling trees.
Findings
Exact domination numbers for hypertrees and sibling trees
Exact total domination numbers for hypertrees and sibling trees
Exact locating-domination and locating-total domination numbers for these structures
Abstract
A locating-dominating set (LDS) of a graph is a dominating set of such that for every two vertices and in , . The locating-domination number is the minimum cardinality of a LDS of . Further if is a total dominating set then is called a locating-total dominating set. In this paper we determine the domination, total domination, locating-domination and locating-total domination numbers for hypertrees and sibling trees.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Domination Parameters in Hypertrees and Sibling trees
Indra Rajasingh1, R. Jayagopal2, and R. Sundara Rajan3,
1,2School of Advanced Sciences, Vellore Institute of Technology,
Chennai - 600 127, India
3Department of Mathematics, Hindustan Institute of Technology and Science,
Chennai - 603 103, India
1[email protected], 2[email protected] and 3[email protected] Corresponding Author: R. JayagopalThis work is supported by National Board of Higher Mathematics (NBHM), No. 2/48(4)/2016/NBHM-RD-II/11580, Department of Atomic Energy (DAE), Government of India.
Abstract
A locating-dominating set (LDS) of a graph is a dominating set of such that for every two vertices and in , . The locating-domination number is the minimum cardinality of a LDS of . Further if is a total dominating set then is called a locating-total dominating set. In this paper we determine the domination, total domination, locating-domination and locating-total domination numbers for hypertrees and sibling trees.
Keywords : Dominating set; total dominating set; locating-dominating set; locating-total dominating set; hypertree; sibling tree.
1 Introduction
A set of vertices in a graph is called a dominating set of if every vertex in is adjacent to some vertex in . The set is said to be a total dominating set of if every vertex in is adjacent to some vertex in . The minimum cardinalities of a dominating set and a total dominating set of are denoted as and , respectively. Domination arises in facility location problems, where the number of facilities such as hospitals or fire stations are fixed and one attempts to minimize the distance that a person needs to travel to get to the closest facility.
Total domination plays a role in the problem of placing monitoring devices in a system in such a way that every site in the system, including the monitors, is adjacent to a monitor site so that, if a monitor goes down, then an adjacent monitor can still protect the system. Installing minimum number of expensive sensors in the system which will transmit a signal at the detection of faults and uniquely determining the location of the faults motivate the concept of locating-dominating sets and locating-total dominating sets [1].
In a parallel computer, the processors and interconnection networks are modeled by the graph , where each processor is associated with a vertex of and a direct communication link between two processors is indicated by the existence of an edge between the associated vertices. Suppose we have limited resources such as disks, input-output connections, or software modules, and we want to place a minimum number of these resource units at the processors, so that every processor is adjacent to at least one resource unit, then finding such a placement involves constructing a minimum dominating set for the graph . Determining if an arbitrary graph has a dominating and locating-dominating sets of a given size are well-known -complete problems [2, 3]. Occurrence of faulty nodes in a device is inevitable. So, to diagnose these faults we make use of locating-total domination set in this system. We place monitoring devices in a system in such a way that every site in the system (including the monitors) is adjacent to a monitor site.
A locating-dominating set in a connected graph is a dominating set of such that for every pair of vertices and in , . The minimum cardinality of a locating-dominating set of is called the locating-domination number [1]. The locating-domination problem has been discussed for paths and cycles [4, 5], infinite grids [6], circulant graphs [7], fault-tolerant graphs [8] and so on.
A locating-total dominating set in a connected graph is a total dominating set of such that for every pair of vertices and in , . The minimum cardinality of a locating total-dominating set of is called the locating-total domination number [1]. The locating-total domination problem has been discussed for trees [9], cubic graphs and grid graphs [10], corona and composition of graphs [11], claw-free cubic graphs [12], edge-critical graphs [13] and so on.
Tree structures are expansible in a natural way, and even unbalanced trees still retain most of the properties that make the tree attractive. Additional links, however, are required to reduce the average distance between nodes and to provide a more uniform message density in all links. An extensive search for the optimal placement of these additional links has shown the half-ring binary trees such as hypertrees, sibling trees and christmas trees to be attractive contenders, primarily because of their simple routing algorithms.
In this paper, we determine the domination, total domination, locating-domination and locating-total domination numbers for hypertrees and sibling trees.
2 Domination in hypertrees
The basic skeleton of a hypertree is a complete binary tree of height . Here the nodes of the tree are numbered as follows: The root node has label . The root is said to be at level [math]. Labels of left and right children are formed by appending a [math] and , respectively to the labels of the parent node. The decimal and binary labels of the hypertree are given in Figure 1. Here the children of the node are labeled as and . Additional links in a hypertree are horizontal and two nodes are joined in the same level of the tree if their label difference is . We denote an -level hypertree as . It has vertices and edges. Hypertree is a multiprocessor interconnection topology which has a frequent data exchange in algorithms such as sorting and Fast Fourier Transforms [14]. The root-fault hypertree , , is a graph obtained from by deleting the root vertex [15]. See Figure 2. The following lemma is obvious from the definition of a hypertree.
Lemma 2.1**.**
The hypertree , , contains disjoint isomorphic copies of and disjoint isomorphic copies of .
Lemma 2.2**.**
Let be the root-fault hypertree . Then .
Proof.
Let be a dominating set of . We claim that . Suppose not, let . Then there exists a vertex in such that deg() = , a contradiction, since . Hence . Let where deg() = deg() = . See Figure 3(a). Now, and hence . Since and are adjacent in , is also a minimum total dominating set of . Therefore . ∎
Lemma 2.3**.**
Let be the root-fault hypertree . Then .
Proof.
Let be a locating-dominating set of . We claim that . By Lemma 2.2, . Assume that . Let where deg() = deg() = . Then and . See Figure 3(b). This implies . Suppose then . Thus . Now let . Then and . Hence . Since vertices and induce a path on 3 vertices in , is also a minimum locating-total dominating set of . Therefore . ∎
Lemma 2.4**.**
Let be the hypertree . Then any minimum dominating set of contains at least vertices from levels and .
Proof.
Let be a minimum dominating set of . Vertices in levels and of induce copies of , each isomorphic to . The worst case arises when both vertices of degree 3 in are already dominated by vertices from . By proof of Lemma 2.2, each copy contains at least 2 vertices of . Hence contains at least vertices from levels and in . ∎
Theorem 2.5**.**
*Let be the hypertree . Then
[TABLE]
Proof.
We prove the result by induction on .
**Case :
**Let and let be a dominating set of . By Lemma 2.4, we need at least 4 vertices from levels 3 and 2 in . To dominate the root vertex, we need at least one vertex from level 1 in or the root vertex itself has to be included in . Therefore . Now we will prove the equality. Let be the set of all vertices comprising of all vertices in level 2 and the root vertex of . See Figure 4. Since all the vertices of level 3 and level 1 are adjacent to the vertices of level 2, is a dominating set of . Therefore . Assume that the result is true for , . That is, Consider . By Lemma 2.1, there are vertex disjoint copies of in . Deletion of these subgraphs along with the vertices of adjacent to vertices of these subgraphs results in . Therefore by Lemma 2.2, and by induction hypothesis, . Now, let and be the minimum dominating sets of and , respectively. Let be the vertex set which contains the vertices from the last three levels of . Similar to the argument for , any minimum dominating set contains at least vertices from levels and . Therefore, \gamma(HT(3k+3))\geq\left|S_{1}\right|+\left|S_{2}\right|=\big{[}(1/7)(2^{3k+2}+3)\big{]}+\big{[}2^{3k+2}\big{]}=(1/7)(2^{(3k+3)+2}+3).
The case when can be dealt with similarly. ∎
Remark 2.6**.**
*The dominating sets described in Theorem 2.5 for , when do not contain any isolated vertex. Let , . By Lemma 2.4, recursively every set of three levels from the bottom of must contain a minimum number of vertices from the last two levels to dominate all the three levels. Hence no minimum dominating set of contains any vertex from level . So far, the vertices of level 0 and level 1 are not dominated. Hence any minimum dominating set contains at least one vertex either from level 0 or level 1. See Figure 5. Now, to make it as a total dominating set, any minimum total dominating set of must include one more vertex either from level [math] or level . See Figure 5. These observations yield the following result. *
Theorem 2.7**.**
*Let be the hypertree . Then
[TABLE]
Lemma 2.8**.**
Let be the root-fault hypertree . Then .
Proof.
Let be a locating-dominating set of . Assume that . The vertices and are the only two vertices of degree 3 in . We assume that and do not belong to . It is easy to see that the removal of and disconnects into two components and which are isomorphic to . See Figure 3(c). We need at least 3 vertices each to identify all the vertices in and . This contradicts the cardinality of . Suppose and belongs to , then we need at least 2 vertices in each of and to dominate and . This again contradicts the cardinality of . The case when either or belongs to is similar. Therefore . Label the vertices of as in Figure 3(c) and let . It is easy to check that is a locating-dominating set of . Further there are no isolated vertices in the subgraph induced by . Therefore is also a locating-total dominating set of . Hence ∎
Remark 2.9**.**
Let be a dominating set of a graph . A pair of vertices and of is said to be located by if . We also say that locates and . If is a locating-dominating set, then locates every pair of vertices in .
Lemma 2.10**.**
Let be the hypertree . Then any minimum locating-dominating set of contains at least vertices of from level .
Proof.
Let be a minimum locating-dominating set of . In , the vertices of level induce a perfect matching consisting of copies of complete graphs, , say . Suppose be an edge in and , then . Therefore contains at least one vertex from each of . ∎
Theorem 2.11**.**
Let be the hypertree . Then
[TABLE]
Proof.
We prove the result by induction on .
**Case :
**Let and be a minimum locating-dominating set of . Let and be the edges in with vertices in level 4 such that is the parent of and , and is the parent of and . See Figure 6. By Lemma 2.10, we need at least 8 vertices from levels 3 in . If , then we need at least one vertex either from level 3 or level 2 to dominate the vertex . If , then to locate the vertices and we need at least one vertex either from level 3 or from level 2. In either case, for four vertices in level 4, at least one vertex from level 3 or level 2 gets included in . Since there are 16 vertices in level 4, at least 4 more vertices from level 3 and level 2 get included in . However, to dominate the root vertex we need at least one vertex from level 1 in or the root vertex itself has to be included in . Therefore . Now we will prove the equality. Let be the set of all vertices in level 0 and level 2 together with four alternate vertices beginning from left to right and the another 4 alternate vertices beginning from right to left in level 4. See Figure 7(b). Now, each vertex of level are located by its children in level 2. Let and be two vertices in level 3 which belongs to . If and has different parent in level 2, then . If and has same parent in level 2, then , since at least one child of and in level 4 are in . Since contains one vertex from every edge of level 4, is a locating-dominating set of . Therefore . Thus, is a minimum locating-dominating set of and hence Assume that the result is true for , . That is, . Consider . By Lemma 2.1, there are vertex disjoint copies of in . Deletion of these subgraphs along with the vertices of adjacent to vertices of these subgraphs, results in . Therefore by Lemma 2.8, and by induction hypothesis, . Now, let and be the minimum locating-dominating set of and , respectively. Let be the vertex set which contains the vertices from the last four levels of . Similar to the argument for , any minimum locating-dominating set contains at least vertices from level and at least vertices from level and level . Therefore, \gamma^{L}(HT(4k+4))\geq\left|S_{1}\right|+\left|S_{2}\right|=\big{[}(1/5)(2^{4k+2}+1)\big{]}+\big{[}2^{4k+3}+2^{4k+2}\big{]}=(1/5)(2^{(4k+2)+4}+1).
The cases when can be dealt with similarly. ∎
Lemma 2.12**.**
Let be the hypertree . Any minimum locating-total dominating set contains at least vertices from levels and in .
Proof.
Let be a minimum locating-total dominating set of . Vertices in levels and of induce copies of , each isomorphic to . The worst case arises when both vertices of degree 3 in are already located by vertices from . By proof of Lemma 2.3, each copy contains at least 3 vertices of . Hence contains at least vertices from levels and in . ∎
Theorem 2.13**.**
Let be the hypertree . Then
[TABLE]
Proof.
We prove the result by induction on .
**Case :
**Let and let be a locating-total dominating set of . By Lemma 2.12, we need at least 6 vertices from levels 2 and 3 in . To dominate the root vertex, we need at least 1 vertex from level 1. Thus . Let be the set of all vertices comprising of all vertices in level 2, one of the vertex in level 1 and two alternate vertices beginning from left to right in level 3. See Figure 8(a). Now, each vertex of level are located by its children in level 3. Let and be two vertices in level 4 which belongs to . Let and be two vertices in level 4 such that and . If and has same parent, say , in level then since either or . If and has different parent, say and , respectively. In this case there are two chances either and not belonging to or any one of them belonging to . If , then since and . If and , then since and . Also it is easy to see that the vertices in level 1 are located by . Thus is a locating-dominating set of . Therefore . Assume that the result is true for , . That is, Consider . By Lemma 2.1, there are vertex disjoint copies of in . Deletion of these subgraphs along with the vertices of adjacent to vertices of these subgraphs results in . Therefore by Lemma 2.3, and by induction hypothesis, . Now, let and be the minimum locating-total dominating set of and , respectively. Let be the vertex set which contains the vertices from the last three levels of . Similar to the argument for , any minimum locating-total dominating set contains at least vertices from levels and . Therefore, \gamma^{L}_{t}(HT(3k+3))\geq\left|S_{1}\right|+\left|S_{2}\right|=\big{[}(1/7)(3(2^{3k+1})+1)\big{]}+\big{[}2^{3k+2}+2^{3k+1}\big{]}=(1/7)(3(2^{(3k+3)+1})+1).
The case when can be dealt with similarly. For illustration, the locating-total dominating set of is given in Figure 8(b).
The case when is similar with being the minimum locating-total dominating set of as the base case. See Figure 7(a). ∎
3 Domination in Sibling trees
Sibling tree is obtained from the complete binary tree by adding edges (sibling edges) between left and right children of the same parent node. Here the nodes of the sibling tree are numbered as follows: The root node has label . The root is said to be at level [math]. Here the children of the nodes are labeled as and . See Figure 9. We denote an -level sibling tree as . It has vertices and edges. For each , let denote the vertex set in level , with . We call the edges in level as terminal edges and the vertices incident on them as terminal vertices. The following lemma is obvious from the definition of a sibling tree.
Lemma 3.1**.**
Let be the sibling tree . Then .
Proof.
Let be a dominating set of . We claim that . Suppose not, let . Then there exists a vertex in such that deg() = , a contradiction since . Hence . Let . Then and hence . Since the vertices in are adjacent in , is also a minimum total dominating set of . Therefore . ∎
The proof of the following lemma is similar to that of Lemma 2.4 and hence is omitted.
Lemma 3.2**.**
Let be the sibling tree . Then any minimum dominating set of contains at least vertices from levels and .
Using Lemma 3.1 and Lemma 3.2, we will prove the following theorem and the proof is similar to that of Theorem 2.5 and hence is omitted.
Theorem 3.3**.**
*Let be the sibling tree . Then
[TABLE]
Remark 3.4**.**
*The dominating sets described in Theorem 3.3 for , when do not contain any isolated vertex. Let , . By Lemma 3.2, recursively every set of three levels from the bottom of must contain a minimum number of vertices from the last two levels to dominate all the three levels. Hence no minimum dominating set of contains any vertex from level . So far, the vertices of level 0 and level 1 are not dominated. Hence any minimum dominating set contains at least one vertex either from level 0 or level 1. See Figure . Now, to make it as a total dominating set, any minimum total dominating set of must include one more vertex either from level [math] or level . See Figure . These observations yield the following result. *
Theorem 3.5**.**
*Let be the sibling tree . Then
[TABLE]
Lemma 3.6**.**
Let be the sibling tree . Then .
Proof.
Let be a locating-dominating set of . We claim . From each terminal edge , at least one vertex of it should belongs to , otherwise , a contradiction. Thus . The terminal vertices do not dominate the vertices in and Hence . Suppose . Let be the non terminal vertex in . If is in , then the root vertex is not dominated. On the other hand, if is in level 0 or 1, then the other two vertices and in are such that , a contradiction. Hence . ∎
The proof of the following lemma is similar to that of Lemma 2.10 and hence is omitted.
Lemma 3.7**.**
Let be the sibling tree . Then any minimum locating-dominating set of contains at least vertices of from level .
Using Lemma 3.6 and Lemma 3.7, we will prove the following theorem and the proof is similar to that of Theorem 2.11 and hence is omitted.
Theorem 3.8**.**
Let be the sibling tree . Then
[TABLE]
Lemma 3.9**.**
Let be the sibling tree . Then .
Proof.
Let be a locating-total dominating set of . We see that at least one vertex from each terminal edge should belong to in order to locate the vertices in level distinctly. Let and be the selected vertices , one each from the two terminal edges. Since and are isolated vertices, to obtain a total domination, we need at least two vertices and such that . Hence . Let . Then clearly is a locating-total dominating set of . See Figure 11(a). Therefore . ∎
Lemma 3.10**.**
Let be the sibling tree . Any minimum locating-total dominating set of contains at least vertices from levels and .
Proof.
Let be a minimum locating-total dominating set of . The vertices in levels and induce a subgraph consisting of copies of complete graph . The worst case arises when the vertices in level are already located by vertices in . However, every should contain 2 vertices of . ∎
Theorem 3.11**.**
Let be the sibling tree . Then
[TABLE]
Proof.
We prove the result by induction on .
**Case :
**Let and let be a locating-total dominating set of . By Lemma 3.10, we need 8 vertices from levels 3 and 2 in . To dominate the root vertex, we need at least 1 vertex from level 1. Thus . Now we will prove the equality. Let be the set of all vertices comprising of all vertices in level 2 and the alternate vertices beginning from left to right in level 3. See Figure . Let and be two vertices in level 3 which belongs to . If and has same parent, then since either or . If and has different parent, then since one vertex from every edge in level 3 belongs to . Also it is easy to see that the vertices in level 0 and level 1 are located by . Thus is a locating-dominating set of . Therefore . Assume that the result is true for , . Consider . Deletion of vertices in levels and in yields . By induction hypothesis, . There are vertex disjoint copies of in the subgraph induced by vertices in the levels and of Therefore by Lemma 3.9, . Now, let and be the minimum locating-total dominating set of and , respectively. Let be the vertex set which contains the vertices from the last three levels of . Similar to the argument for , any minimum locating-total dominating set contains at least vertices from levels and . Therefore \gamma^{L}_{t}(ST_{3k+3})\geq\left|S_{1}\right|+\left|S_{2}\right|=\big{[}(1/7)(2^{3k+3}-1)\big{]}+\big{[}2^{3k+3}\big{]}=(1/7)(2^{(3k+3)+3}-1).
The case when can be dealt with similarly. ∎
4 Conclusion
In this paper, we have proved that when is a hypertree , and when is . We have also computed and , . We have obtained similar results for sibling tree . Finding classes of graphs with is under investigation.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1[1] T. W. Haynes, M. A. Henning, J. Howard , Locating and total dominating sets in trees , Discrete Applied Mathematics, Vol. 154, no. 8, 1293-1300, 2006.
- 2[2] M. R. Garey and D. S. Johnson, Computers and intractability: A guide to the theory of NP-completeness , W. H. Freeman company publisher, San Francisco, US, 1979.
- 3[3] I. Charon, O. Hudry and A. Lobstein, Minimizing the size of an identifying or locating-dominating code in a graph is NP-hard , Theoretical Computer Science, Vol. 290, no. 3, 2109-2120, 2006.
- 4[4] G. Exoo, Locating-dominating codes in cycles , Australasian Journal of Combinatorics, Vol. 49, 177-194, 2011.
- 5[5] C. Chen, C. Lu and Z. Miao, Identifying codes and locating-dominating sets on paths and cycles , Discrete Applied Mathematics, Vol. 159, no. 15, 1540-1547, 2011.
- 6[6] I. Honkala and T. Laihonen, On locating-dominating sets in infinite grids , European Journal of Combinatorics, Vol. 27, no. 2, 218-227, 2006.
- 7[7] M. Ghebleha and L. Niepelb, Locating and identifying codes in circulant networks , Discrete Applied Mathematics, Vol. 161, no. 13-14, 2001-2007, 2013.
- 8[8] P. J. Slater, Fault-tolerant locating-dominating sets , Discrete Mathematics, Vol. 249, no. 1-3, 179-189, 2002.
