
Resolving set has several applications in the fields of science, engineering, and computer science. One application of the resolving set problem includes navigation robots, chemical structures, and supply chain management. Suppose the set W={s1,s2,…,sk}⊂V(G), the vertex representations of x∈V(G) is rm(x|W)={d(x,s1),d(x,s2),…,d(x,sk)}, where d(x,si) is the length of the shortest path of the vertex x and the vertex in W together with their multiplicity. The set W is called a local m-resolving set of graphs G if rm(v|W)≠rm(u|W) for uv∈E(G). The local m-resolving set having minimum cardinality is called the local multiset basis and its cardinality is called the local multiset dimension of G, denoted by mdl(G). In our paper, we determined the bounds of the local multiset dimension of the comb product of tree graphs.
Citation: Ridho Alfarisi, Liliek Susilowati, Dafik. Local multiset dimension of comb product of tree graphs[J]. AIMS Mathematics, 2023, 8(4): 8349-8364. doi: 10.3934/math.2023421
[1] | Dalal Awadh Alrowaili, Uzma Ahmad, Saira Hameeed, Muhammad Javaid . Graphs with mixed metric dimension three and related algorithms. AIMS Mathematics, 2023, 8(7): 16708-16723. doi: 10.3934/math.2023854 |
[2] | Yuni Listiana, Liliek Susilowati, Slamin Slamin, Fadekemi Janet Osaye . A central local metric dimension on acyclic and grid graph. AIMS Mathematics, 2023, 8(9): 21298-21311. doi: 10.3934/math.20231085 |
[3] | Ahmed Alamer, Hassan Zafar, Muhammad Javaid . Study of modified prism networks via fractional metric dimension. AIMS Mathematics, 2023, 8(5): 10864-10886. doi: 10.3934/math.2023551 |
[4] | Supachoke Isariyapalakul, Witsarut Pho-on, Varanoot Khemmani . The true twin classes-based investigation for connected local dimensions of connected graphs. AIMS Mathematics, 2024, 9(4): 9435-9446. doi: 10.3934/math.2024460 |
[5] | Abdessatar Souissi, El Gheteb Soueidy, Mohamed Rhaima . Clustering property for quantum Markov chains on the comb graph. AIMS Mathematics, 2023, 8(4): 7865-7880. doi: 10.3934/math.2023396 |
[6] | Hashem Bordbar, Sanja Jančič-Rašovič, Irina Cristea . Regular local hyperrings and hyperdomains. AIMS Mathematics, 2022, 7(12): 20767-20780. doi: 10.3934/math.20221138 |
[7] | S. Gajavalli, A. Berin Greeni . On strong geodeticity in the lexicographic product of graphs. AIMS Mathematics, 2024, 9(8): 20367-20389. doi: 10.3934/math.2024991 |
[8] | Xia Hong, Wei Feng . Completely independent spanning trees in some Cartesian product graphs. AIMS Mathematics, 2023, 8(7): 16127-16136. doi: 10.3934/math.2023823 |
[9] | Chenggang Huo, Humera Bashir, Zohaib Zahid, Yu Ming Chu . On the 2-metric resolvability of graphs. AIMS Mathematics, 2020, 5(6): 6609-6619. doi: 10.3934/math.2020425 |
[10] | Ali N. A. Koam . Metric based resolvability of cycle related graphs. AIMS Mathematics, 2024, 9(4): 9911-9925. doi: 10.3934/math.2024485 |
Resolving set has several applications in the fields of science, engineering, and computer science. One application of the resolving set problem includes navigation robots, chemical structures, and supply chain management. Suppose the set W={s1,s2,…,sk}⊂V(G), the vertex representations of x∈V(G) is rm(x|W)={d(x,s1),d(x,s2),…,d(x,sk)}, where d(x,si) is the length of the shortest path of the vertex x and the vertex in W together with their multiplicity. The set W is called a local m-resolving set of graphs G if rm(v|W)≠rm(u|W) for uv∈E(G). The local m-resolving set having minimum cardinality is called the local multiset basis and its cardinality is called the local multiset dimension of G, denoted by mdl(G). In our paper, we determined the bounds of the local multiset dimension of the comb product of tree graphs.
One of the topics in distances in graphs is the resolving set problem. This topic has many applications in science and technology namely the application of resolving set problems can be found in network infrastructure, navigation robots, chemistry structures, and computer science. The application of metric dimension in networks is one of the described navigation robots. Each place is called the vertex and the connections between vertex are called edges. The minimum number of robots is required for each location and the vertex of some networks is called resolving set problems, for more detail this application is in [1].
All graphs G are simple and connected graphs. We have the vertex set and edge set, respectively are V(G) and E(G). The distance of u and v and denoted by d(u,v) is the length of the shortest path of the vertices u to v. For the set W={s1,s2,...,sk}⊂V(G). The vertex representations of the vertex x to the set W is an ordered k-tuple, r(x|W)=(d(x,s1),d(x,s2),...,d(x,sk)). The set W is called the resolving set of G if every vertex of G has different vertex representations. The resolving set having minimum cardinality is called basis and its cardinality is called the metric dimension of G and denoted by dim(G) by [2]. Okamoto et al [3] introduced a new variant of resolving set problems which are called local resolving set problems. In his paper its concept is called local multiset dimension of graphs G. The set W is called a local resolving set if ∀xy∈E(G),r(x|W)≠r(y|W). The local resolving set having minimum cardinality is called local basis and its cardinality is called the local metric dimension of G and denoted by ldim(G).
Simanjuntak et al. [4] introduced the multiset dimension of graphs G. Suppose the set ={s1,s2,…,sk}⊂V(G), the vertex representations of a vertex x∈V(G) to the set W is the multiset, rm(x|W)={d(x,s1),d(x,s2),…,d(x,sk)} where d(x,si) is the length of the shortest path of the vertex x and the vertex in W together with their multiplicities. The set W is called an m-resolving set if ∀xy∈E(G),rm(x|W)≠rm(y|W). If G has an m-resolving set, then an m-resolving set having minimum cardinality is called a multiset basis and its cardinality is called the multiset dimension of graphs G and denoted by md(G) and we say that G has an infinite multiset dimension and we write md(G)=∞.
Alfarisi et al. [5] defined a new notion based on the multiset dimension of G, namely a local multiset dimension. Suppose the set ={s1,s2,…,sk}⊂V(G), the vertex representations of a vertex x∈V(G) to the set W is rm(x|W)={d(x,s1),d(x,s2),…,d(x,sk)}. The set W is called a local m-resolving set of G if rm(v|W)≠rm(u|W) for uv∈E(G). The local m-resolving set having minimum cardinality is called the local multiset basis and its cardinality is called the local multiset dimension and denoted by mdl(G) and we say that G has an infinite local multiset dimension and we write mdl(G)=∞. Alfarisi et al. [6] determined multiset dimension problems of almost hypercube graphs.
We have some results on the local multiset dimension of some known graphs namely path, star, tree, and cycle and also the local multiset dimension of graph operations namely, cartesian product [6], m-shadow graph [7], and some related cycles [8]. Adawiyah et al. [9] also studied the local multiset dimension of unicyclic graphs. There are some results used for proving the other results as follows.
Lemma 1. [10] Let G be a connected graph and W⊂V(G). If W contains a resolving set of G, then W is a resolving set of G.
Proposition 1. [11] A graph is bipartite if and only if it contains no odd cycle.
Proposition 2. [12] The local multiset dimension of G is one if and only if G is a bipartite graph.
Proposition 3. [12] If T is a tree graph with order n, then mdl(T)=1.
Definition 1. [13] Let G and H be two connected graphs. Let o be a vertex of H. The comb product of G and H, denoted by G⊳H, is a graph obtained by taking one copy of G and |V(G)| copies of H and identifying the i−th copy of H at the vertex o with the i−th vertex of G.
Lemma 2. Let G and H be a connected graph. Graph G⊳H is a bipartite graph if and only if G and H is a bipartite graph.
In this section, we investigated the local multiset dimension of the graph resulting of the comb product of tree and cycle. We have determined the bounds and the exact value of the local multiset dimension of the comb product of the tree and cycle.
Lemma 3. Let T be a tree graph and Cm be a cycle graph for m≥4. Then
mdl(T⊳Cm)=1 , for m even, |
mdl(T⊳Cm)≥n, for m odd. |
Proof. The comb product of tree graph T with order n and cycle Cm for m≥3, denoted by T⊳Cm. This graph has a backbone and leaves where a backbone is tree graph T and leaves are the subgraph cycle (cycle leaves) such that we have n-cycle leaves. The graph T⊳Cm has V(T⊳Cm)={vi,j;i∈[1,n] and j∈[1,m]} and E(T⊳Cm)=E(T)∪⋃i=ni=1(Cm)i where (Cm)i is a i-th cycle leaves. The vertex in the backbone is called the terminal vertex and the vertex in cycle leaves is called a leaves vertex. From Figure 2, that vi,1 is terminal vertex and vi,j(j≠1) is leaves vertex.
Case 1. For m is even
A cycle graph Cm with m even is a bipartite graph. Based on Lemma 2 that T⊳Cm is a bipartite graph. Since T⊳Cm is a bipartite graph, based on Proposition 2, mdl(T⊳Cm)=1.
Case 2. For m is odd
Based on Lemma 2 that T⊳Cm isn't a bipartite graph, such that mdl(T⊳Cm)>1. We prove that mdl(T⊳Cm)≥n. We will prove that mdl(T⊳Cm)≥n. Taking any P⊂V(T⊳Cm) with |P|=n−1. Graph T⊳Cm has n copies of Cm, namely (Cm)1, (Cm)2, …, (Cm)n. Thus, we know that there is at least one cycle that hasn't been resolver. Since (Cm)k for 1≤k≤n isn't contained resolver, such that the two adjacent vertices vk,m+12,vk,m+32 in (Cm)k have the same distance to the terminal vertex, d(vk,m+12,vk,1)=m+12−1=m−12 and d(vk,m+32,vk,1)=m+1−m+32=m−12. We know that
d(vk,m+12,vi,m)=d(vk,m+12,vk,1)+d(vk,1,vi,m)=m−12+d(vk,1,vi,m), | (1) |
d(vk,m+32,vi,m)=d(vk,m+32,vk,1)+d(vk,1,vi,m)=m−12+d(vk,1,vi,m). | (2) |
It is clear that d(vk,m+12,vi,m)=d(vk,m+32,vi,m) such that rm(vk,m+12|W)=rm(vk,m+32|W). Hence, mdl(T⊳Cm)≥n.
Theorem 1. Let Pn⊳Cm be a comb product of path and cycle for n≥2 and m≥3. Then
mdl(Pn⊳Cm)={1,for m is even ,n,for m is odd and m ≠3 or for m=3 and n is odd,n+1,for m=3 and n is even. |
Proof. The graphs Pn⊳Cm has V(Pn⊳Cm)={vi,j;i∈[1,n]andj∈[1,m]} and E(Pn⊳Cm)={vi,1vi+1,1;i∈[1,n−1]}∪{vi,jvi,j+1};i∈[1,n],j∈[1,m−1]}∪{vi,mvi,1;i∈[1,n]} where (Cm)i is a i-th cycle leaves. The vertex vi,1 in the backbone is called terminal vertex and vertex vi,j(j≠1) in cycle leaves is called a leaves vertex. This proof is divided into four cases as follows.
Case 1. For m is even
A cycle graph Cm with m even is a bipartite graph. Based on Lemma 2 that Pn⊳Cm is a bipartite graph. Since Pn⊳Cm is a bipartite graph, based on Proposition 2, mdl(Pn⊳Cm)=1.
Case 2. For m is odd and m≠3
Choose W={v1,m−1,vi,m;i∈[2,n]}, so that |W|=n. We are going to prove that the vertex representations of two adjacent vertices in Pn⊳Cm are distinct. The resolver vertex in (Cm)i,i∈[1,n] exactly one resolver in every cycle leaves. The resolver v1,m−1 in (Cm)1 and vi,m in (Cm)i for i∈[2,n]. In the first step, we show that vertex representations of two adjacent vertices in (Cm)i,i∈[2,n] respect to W are distinct as follows
(1) The number of leave vertex in every cycle leave is even, such that two adjacent vertices in cycle leaves have the same distance to the terminal vertex. The vertex vi,m+12,vi,,m+32 in (Cm)i so that, d(vi,m+12,vi,1)=d(vi,,m+32,vi,1).
(2) The number of leave vertex in every cycle leave is even, such that two adjacent vertices in cycle leaves have the same distance to the resolver. The vertex vi,m−12,vi,m+12 in (Cm)i so that, d(vi,m−12,vi,m)=d(vi,m+12,vi,m).
(3) We are going to show that two adjacent vertices in cycle leaves have distinct vertex representations with respect to W. For vk,m∈W,k∈[2,n] and k≠i, we know that d(vi,m−12,vi,1)≠d(vi,m+12,vi,1) such that
d(vi,m−12,vk,m)=d(vi,m−12,vi,1)+d(vi,1,vk,m), | (3) |
d(vi,m+12,vk,m)=d(vi,m+12,vi,1)+d(vi,1,vk,m). | (4) |
Therefore, d(vi,m−12,vk,m)≠dd(vi,m+12,vk,m) so that rm(vi,m−12|W)≠rm(vi,m+12|W).
(4) Take two adjacent vertices, vi,r,vi,s∈V((Cm)i)−{vi,m−12,vi,m+12};r,s∈[1,m],r≠s. The resolver vk,m∈W,k∈[2,n] and k≠i, we know that d(vi,r,vi,1)≠d(vi,s,vi,1) such that
d(vi,r,vk,m)=d(vi,r,vi,1)+d(vi,1,vk,m), | (5) |
d(vi,s,vk,m)=d(vi,s,vi,1)+d(vi,1,vk,m). | (6) |
Therefore, d(vi,r,vk,m)≠d(vi,s,vk,m) so that rm(vi,r|W)≠rm(vi,s|W).
(5) The backbone indices path Pn,n odd have a symmetry vertex namely vr,1andvs,1 with r+s=n+1 and vr,1 not adjacent to vs,1 because there one vertex vn+12 as center vertex.
(6) The terminal vertex indices path graph such that the distance between two adjacent terminal vertex to the resolver vertex is distinct, d(vk,1,vi,m)≠d(vl,1,vi,m) and d(vk,1,v1,m−1)≠d(vl,1,v1,m−1) for k,l∈[2,n],k≠l≠i. Thus, rm(vk,1|W)≠rm(vl,1|W).
(7) The vertex vi,1 adjacent to vi,2, d(vi,1,vi,m)≠d(vi,2,vi,m) and
d(vi,1,vk,m)=d(vi,1,vi,1)+d(vi,1,vk,m)=d(vi,1,vk,m), | (7) |
d(vi,2,vk,m)=d(vi,2,vi,1)+d(vi,1,vk,m)=1+d(vi,1,vk,m),k≠i. | (8) |
It is clear that d(vi,1,vk,m)≠d(vi,2,vk,m), we know that rm(vi,1|W)={d(vi,1,vi,m),d(vi,1,vk,m)}≠ {d(vi,2,vi,m),d(vi,2,vk,m)=rm(vi,2|W)}.
Now, we prove that vertex representation of two adjacent vertices in (Cm)1 respect to W is distinct. For v1,m−1 and vk,m∈W,k∈[2,n] and k≠i, we know that d(v1,m−32,v1,m−1)=d(v1,m−12,v1,m−1) and d(v1,m−32,v1,1)≠d(v1,m−12,v1,1) such that
d(v1,m−32,vk,m)=d(v1,m−32,v1,1)+d(v1,1,vk,m), | (9) |
d(v1,m−12,vk,m)=d(v1,m−12,v1,1)+d(v1,1,vk,m). | (10) |
It is clear that d(v1,m−32,vk,m)≠d(v1,m−12,vk,m) so that rm(v1,m−12|W)≠rm(v1,m+12|W). Next, taking two adjacent vertices, v1,r,v1,s∈V((Cm)1)−{vv1,m−32,v1,m−12};r,s∈[1,m],r≠s. We know that d(v1,r,v1,m−1)=d(v1,s,v1,m−1) and d(v1,r,v1,1)≠d(v1,s,v1,1) such that
d(v1,r,vk,m)=d(v1,r,v1,1)+d(v1,1,vk,m), | (11) |
d(v1,s,vk,m)=d(v1,s,v1,1)+d(v1,1,vk,m). | (12) |
Therefore, d(v1,r,vk,m)≠d(v1,s,vk,m) so that rm(v1,r|W)≠rm(v1,s|W). Hence, rm(vi,j|W)≠rm(vk,l|W) for vi,j adjacent to vk,l for i∈[1,n]. Consequently, W is a local m-resolving set of Pn⊳Cm. Based on Lemma 3 that mdl(Pn⊳Cm)≥n. Thus, mdl(Pn⊳Cm)=n. This completes the proof.
Case 3. For m=3 and n is even
We choose W={vi,3;i∈[1,n]}∪{vn,1}, so that |W|=n+1. The resolver vertex in (C3)i,i∈[1,n] exactly one resolver in every cycle leaves and one resolver in the terminal vertex. We are going to prove that the vertex representations of two adjacent vertices in Pn⊳C3 are distinct. We show that vertex representations of two adjacent vertices in (Cm)i,i∈[2,n] with respect to W are distinct.
(1) In the first step, we focus on resolver in cycle leaves (vi,3), we have rm(vn2,1|W−{vn,1})=rm(vn+22,1|W−{vn,1}). Because n even, there are a symmetry vertex namely vk,1 and vl,1 with k+l=n+1 and k≠l. We know that vn2,1,vl,n+22 as two adjacent vertices and symmetry vertex in Pn⊳C3 such that rm(vn2,1|W−{vn,1})={1,22,32,…,(n2)2,n+22}=rm(vn+22,1|W−vn,1).
(2) Now, d(vn2,1,vn,1)=n−n2=n2 and d(vn+22,1,vn,1)=n−n+22=n−22 so that d(vn2,1,vn,1)≠d(vn+22,1,vn,1).
Based on points (1) and (2) that
rm(vn2,1|W)={1,22,32,…,(n2)2,n+22,d(vn2,1,vn,1)}, | (13) |
rm(vn+22,1|W)={1,22,32,…,(n2)2,n+22,d(vn+22,1,vn,1)}. | (14) |
It is clear that rm(vn2,1|W)≠rm(vn+22,1|W) for vn2,1 adjacent to vn+22,1. Consequently, W is a local m-resolving set of Pn⊳C3.
Furthermore, we prove that W is the local m-resolving set with minimum cardinality. Taking any set S⊂V(Pn⊳C3) with |S|<|W|. Let |S|=n,
(1) u∈W, every vertex u in cycle leave.
Every cycle leave has one vertex as a resolver. There are the two adjacent terminal vertices in the backbone indices path Pn. Because n even, then there is a symmetry vertices namely vk,1 and vl,1 with k+l=n+1 and k≠l. We know that vn2,1,vl,n+22 as two adjacent vertices and symmetry vertices in Pn⊳C3 such that rm(vn2,1|W)={1,22,32,…,(n2)2,n+22}=rm(vn+22,1|W).
(2) u∈W, there is at least one resolver that isn't in cycle leaves.
If there is at least one resolver that isn't in cycle leaves, then there are the two adjacent vertices in cycle leaves that have the same vertex representations. We take (C3)n without resolver such that
d(vn,2,vn,1)=d(vn,3,vn,1)=1, | (15) |
d(vn,2,vk,3)=d(vn,2,vn,1)+d(vn,1,vk,3), | (16) |
d(vn,3,vk,3)=d(vn,3,vn,1)+d(vn,1,vk,3), | (17) |
d(vn,2,y)=d(vn,2,vn,1)+d(vn,1,y);y∈W,y in vi,1 or cycle leaves, | (18) |
d(vn,3,y)=d(vn,3,vn,1)+d(vn,1,y);y∈W,y in vi,1 or cycle leaves. | (19) |
Based on above cases that rm(vn,2|W)={d(vn,2,vk,3),d(vn,2,y)}={d(vn,3,vk,3),d(vn,3,y)}=rm(vn,3|W). Therefore, S is not a local m-resolving set of Pn⊳C3. Thus, mdl(Pn⊳C3)=n+1. This completes the proof.
Theorem 2. Let T1 and T2 be tree graphs, then mdl(T1⊳T2)=1.
Proof. Based on Lemma 2 that T1⊳T2 is a bipartite graph. Since T1⊳T2 is a bipartite graph, based on Proposition 2, mdl(T1⊳T2)=1.
Theorem 3. Let T be a tree graph and Cn be a cycle graph for n≥3. Then
mdl(Cn⊳T)={1,n even,2,n odd. |
Proof. The graph Cn⊳T has V(Cn⊳T)=⋃i=ni=1V(Ti) and E(Cn⊳T)={v1,1vn,1,vi,1vi+1,1;i∈[1,n−1]}∪⋃i=ni=1E(Ti) where Ti is a i-th tree leaves. We choose a vertex ai∈Ti. The vertex vi,1 in the backbone is called terminal vertex and vertex vi,j(j≠1) in tree leaves is called a leaves vertex. This proof is divided into four cases as follows.
Case 1. For n is even
A cycle graph Cn with m even is a bipartite graph. Based on Lemma 2 that Cn⊳T is a bipartite graph. Since Cn⊳T is a bipartite graph, based on Proposition 2, mdl(Cn⊳T)=1.
Case 2. For n is odd
Choose W={v1,1,a2} with a2∉V(Cn) or a2∈V(T2), so that |W|=2. We are going to prove that the vertex representations of two adjacent vertices in Cn⊳T are distinct. We can see that
rm(vi,1|W)={i−1,d(vi,1,v2,1)+dT2(v2,1,a2)},i∈[1,n2+1], | (20) |
rm(vi,1|W)={m−i+1,d(vi,1,v2,1)+dT2(v2,1,a2)},i∈[n2+2,n], | (21) |
rm(a1|W)={dT1(a1,vi,1),1+dT2(v2,1,a2)}, | (22) |
rm(a2|W)={dT2(a2,v2,1)+1,0}, | (23) |
rm(a′2|W)={dT2(a2,v2,1)+1,dT2(a′2,a2)},a′2∉W, | (24) |
rm(ai|W)={dTi(ai,vi,1)+i−1,dTi(ai,vi,1)+dT2(v2,1,a2)+i−2},i∈[3,n+12], | (25) |
rm(ai|W)={dTi(ai,vi,1)+m−i+1,dTi(ai,vi,1)+d(vi,1,v2,1)+dT2(v2,1,a2)}, i∈[n+12+1,n]. | (26) |
The vertex representations of the vertices vi,1 and ai are distinct and so W is a local m-resolving set of Cn⊳T. Thus, mdl(Cn⊳T)≤2. It concluded that mdl(Cn⊳T)=2.
Furthermore, we will prove that W is the local m-resolving set with minimum cardinality. Take any set S⊂V(Cn⊳T) with |S|<|W|. Let |S|=1, suppose the local m−resolving set S={v} so that there are some conditions of this proof as follows
(1) If v∈V(Cn) namely v=vi,1, then
rm(v(n+12+i−1)mod n,1|W)=rm(v(n+32+i−1)mod n,1|W)={n−12}, | (27) |
(2) If v∈V(Ti), then
rm(v(n+12+i−1)mod n,1|W)=rm(v(n+32+i−1)mod n,1|W)={n−12+dTi(vi,1,v)}. | (28) |
It is clear that rm(v(n+12+i−1)mod n, 1|W)=rm(v(n+32+i−1)mod n, 1|W). Therefore, S is not a local m-resolving set of Cn⊳T. Thus, mdl(Cn⊳T)=2.
Theorem 4. Let Cn and Cm be cycle graphs for n,m≥3. Then
mdl(Cn⊳Cm)={1,both n and m are even,2,one of the n and m is even and the other is odd,n,both n and m are odd. |
Proof. The graph Cn⊳Cm has V(Cn⊳Cm)={vi,j;i∈[1,n],j∈[1,m]} and E(Cn⊳Cm)={v1,1vn,1,vi,1vi+!,1;i∈[1,n−1]}∪{vi,jvi,j+1};i∈[1,n],j∈[1,m−1]}∪{vi,mvi,1;i∈[1,n]} where (Cm)i is an i-th cycle leaves. The vertex vi,1 in the backbone is called terminal vertex and vertex vi,j(j≠1) in cycle leaves is called a leaves vertex. This proof is divided into four cases as follows.
Case 1. For n and m are even
A cycle graph Cn and Cm with n,m even is a bipartite graph. Based on Lemma 2 that Cn⊳Cm is a bipartite graph. Since Cn⊳Cm is a bipartite graph, based on Proposition 2, mdl(Cn⊳Cm)=1.
Case 2. For n is odd and m is even
Choose W={v1,1,v2,m}, so that |W|=2. We are going to prove that the vertex representations of two adjacent vertices in Cn⊳Cm are distinct. We can see that:
Vertices v | Representation rm(v|W) | Conditions |
v1,j | {j−1,j+1} | j∈[1,m2+1] |
v1,j | {m−j+1,m−j+3} | j∈[m2+2,m] |
v2,j | {j,j} | j∈[1,m2] |
v2,j | {m−j,m−j+2} | j∈[m2+1,m] |
vi,j | {i+j−2,i+j−2} | i∈[3,n+12],j∈[1,m2+1] |
vi,j | {m−j+i,m−j+i} | i∈[3,n+12],j∈[m2+2,m] |
vi,j | {n−i+j,i+j−2} | i=n+32,j∈[1,m2+1] |
vi,j | {m+n−j−i+2,m−j+i} | i=n+32,j∈[m2+2,m] |
vi,j | {n−i+j,n−i+j+2} | i∈[n+52,n],j∈[1,m2+1] |
vi,j | {m+n−j−i+2,m+n−j−i+4} | i∈[n+52,n],j∈[m2+2,m] |
The vertex representations of the adjacent vertices vi are distinct such that W is a local m-resolving set of Cn⊳Cm. Thus, mdl(Cn⊳Cm)≤2.
Furthermore, we are going to prove that W is the local m-resolving set with minimum cardinality. Take any set S⊂V(Cn⊳Cm) with |S|<|W|. Let |S|=1, suppose the local m-resolving set W={v} so that there are some conditions of this proof as follows
(1) If v∈V(Cn) namely v=vi,1, then we have vertex representation as follows
rm(v(n+12+i−1)mod n,1|W)=rm(v(n+32+i−1)mod n,1|W)={n−12}. | (29) |
(2) If v∈V((Cm)i), then we have vertex representation as follows
rm(v(n+12+i−1)mod n,1|W)=rm(v(n+32+i−1)mod n,1|W)={n−12+d(vi,1,v)}. | (30) |
It is clear that rm(v(n+12+i−1)mod n, 1|W)=rm(v(n+32+i−1)mod n, 1|W). Therefore, S is not a local m-resolving set of Cn⊳Cm. Thus, mdl(Cn⊳Cm)=2.
Case 3. For n is even and m is odd
Choose W={v1,m−1,vi,m;i∈[2,n]}, so that |W|=n. We are going to prove that the vertex representations of two adjacent vertices in Cn⊳Cm are distinct. The resolver vertex in (Cm)i,i∈[1,n] exactly one resolver in every cycle leaves. The resolver v1,m−1 in (Cm)1 and vi,m in (Cm)i for i∈[2,n]. In the first step, we show that vertex representations of two adjacent vertices in (Cm)i,i∈[2,n] with respect to W are distinct as follows
(1) The number of leave vertex in every cycle leave is even, such that two adjacent vertices in cycle leaves have the same distance to the terminal vertex. The vertex vi,m+12,vi,m+32 in (Cm)i so that, d(vi,m+12},vi,1)=d(vi,m+32,vi,1).
(2) The number of leave vertex in every cycle leave is even, such that two adjacent vertices in cycle leaves have the same distance to the resolver. The vertex vi,m−12,vi,m+12 in (Cm)i so that, d(vi,m−12,vi,m)=d(vi,m+12,vi,m).
(3) We are going to show that two adjacent vertices in cycle leaves have distinct vertex representations with respect to W. For vk,m∈W,k∈[2,n] and k≠i, we know that d(vi,m−12,vi,1)≠d(vi,m+12,vi,1) such that
d(vi,m−12,vk,m)=d(vi,m−12,vi,1)+d(vi,1,vk,m), | (31) |
d(vi,m+12,vk,m)=d(vi,m+12,vi,1)+d(vi,1,vk,m). | (32) |
Therefore, d(vi,m−12,vk,m)≠d(vi,m+12,vk,m) so that rm(vi,m−12|W)≠rm(vi,m+12|W).
(4) Take two adjacent vertices, vi,r,vi,s∈V((Cm)i)−{vi,m−12,vi,m+12},r,s∈[1,m],r≠s. The resolver vk,m∈W, k∈[2,n] and k≠i, we know that d(vi,r,vi,s)≠d(vi,s,vi,1) such that
d(vi,r,vk,m)=d(vi,r,vi,1)+d(vi,1,vk,m), | (33) |
d(vi,s,vk,m)=d(vi,s,vi,1)+d(vi,1,vk,m). | (34) |
Therefore, d(vi,r,vk,m)≠d(vi,s,vk,m) so that rm(vi,r|W)≠rm(vi,s|W).
(5) The terminal vertex indices cycle graph Cn,n is odd such that the distance between two adjacent terminal vertex to the resolver vertex is distinct, d(vk,1,vi,m)≠d(vl.1,vi,m) and d(vk,1,v1,m−1)≠d(vl,1,v1,m−1) for k,l∈[2,n],k≠l≠i. Based on (6) so that rm(vk,1|W)≠rm(vl,1|W).
(6) The vertex vi,1 adjacent to vi,2,d(vi,1,vi,m)≠d(vi,2,vi,m) and
d(vi,1,vk,m)=d(vi,1,vi,1)+d(vi,1,vk,m)=d(vi,1,vk,m), | (35) |
d(vi,2,vk,m)=d(vi,2,vi,1)+d(vi,1,vk,m)=1+d(vi,1,vk,m),k≠i. | (36) |
It is clear that d(vi,1,vk,m)≠d(vi,2,vk,m), we know that rm(vi,1|W)={d(vi,1,vi,m),d(vi,1,vk,m)}≠{d(vi,2,vi,m),d(vi,2,vk,m)}=rm(vi,2|W).
Now, we prove that vertex representation of two adjacent vertices in (Cm)1 respect to W is distinct. For v1,m−1 and vk,m∈W,k∈[2,n] and k≠i, we know that d(v1,m−32,v1,m−1)=d(v1,m−12,v1,m−1) and d(v1,m−32,v1,1)≠d(v1,m−12,v1,1) such that
d(v1,m−32,vk,m)=d(v1,m−32,v1,1)+d(v1,1,vk,m), | (37) |
d(v1,m−12,vk,m)=d(v1,m−12,v1,1)+d(v1,1,vk,m). | (38) |
It is clear that d(v1,m−32,vk,m)≠d(v1,m−12,vk,m) so that rm(v1,m−12|W)≠rm(v1,m+12|W). Next, taking two adjacent vertices, v1,r,v1,s∈V((Cm)1)−{v1,m−32,v1,m−12},r,s∈[1,m],r≠s. We know that d(v1,r,v1,m−1)=d(v1,s,v1,m−1) and d(v1,r,v1,1)≠d(v1,s,v1,1) such that
d(v1,r,vk,m)=d(v1,r,v1,1)+d(v1,1,vk,m), | (39) |
d(v1,s,vk,m)=d(v1,s,v1,1)+d(v1,1,vk,m). | (40) |
Therefore, d(v1,r,vk,m)≠d(v1,s,vk,m) so that rm(v1,r|W)≠rm(v1,s|W). Hence, rm(vi,j|W)≠rm(vk,l|W) for vi,j adjacent to vk,l for i∈[1,n]. Consequently, W is a local m-resolving set of Cn⊳Cm.
Furthermore, we will prove that W is the local m-resolving set with minimum cardinality. Taking any set S⊂V(Cn⊳Cm) with |S|<|W|. Let |S|=n−1, there are at least one cycle leaves which haven't resolver. For m odd, there are even leave vertex such that the two adjacent vertices vk,m+12,vk,m+32∈V(Cm)k,k∈[1,m] have the same distance to the terminal vertex, d(vk,m+12,vk,1)=m+12−1=m−12 and d(vk,m+32,vk,m)=m+1−m+32=m−12. We know that
d(vk,m+12,vi,m)=d(vk,m+12,vk,1)+d(vk,1,vi,m)=m−12+d(vk,1,vi,m), | (41) |
d(vk,m+32,vi,m)=d(vk,m+32vk,1)+d(vk,1,vi,m)=m−12+d(vk,1,vi,m). | (42) |
It is clear that d(vk,m+12,vi,m)=d(vk,m+32,vi,m) such that rm(vk,m+12|W)=rm(vk,m+32|W). Therefore, S is not a local m-resolving set of Cn⊳Cm. Thus, mdl(Cn⊳Cm)=n. This completes the proof.
Case 4. For n and m are odd
Choose W={v1,m−1,v3,m−1,v4,m−1,vi,m;i∉{1,3,4}}, so that |W|=n. We are going to prove that the vertex representations of two adjacent vertices in Cn⊳Cm are distinct. The resolver vertex in (Cm)i,i∈[1,n] exactly one resolver in every cycle leaves. The resolver vi,m−1 in (Cm)i,i∈{1,3,4} and vi,m in (Cm)i for i∉{1,3,4}. In the first step, we show that vertex representations of two adjacent vertices in (Cm)i,i∉{1,3,4} with respect to W are distinct as follows
(1) The number of leave vertex in every cycle leaves is even, such that two adjacent vertices in cycle leaves have the same distance to the terminal vertex. The vertex vi,m+12,vi,m+32 in (Cm)i so that, d(vi,m+12,vi,1)=d(vi,m+32,vi,1).
(2) The number of leave vertex in every cycle leaves is even, such that two adjacent vertices in cycle leaves have the same distance to the resolver. The vertex vi,m−12,vi,m+12 in (Cm)i so that, d(vi,m−12,vi,m)=d(vi,m+12,vi,m).
(3) We are going to show that two adjacent vertices in cycle leaves have distinct vertex representations with respect to W. For vk,m∈W,k∈[2,n] and k≠i, we know that d(vi,m−12,vi,1)≠d(vi,m+12,vi,1) such that
d(vi,m−12,vk,m)=d(vi,m−12,vi,1)+d(vi,1,vk,m), | (43) |
d(vi,m+12,vk,m)=d(vi,m+12,vi,1)+d(vi,1,vk,m). | (44) |
Therefore, d(vi,m−12,vk,m≠d(vi,m+12,vk,m) so that rm(vi,m−12|W)≠rm(vi,m+12|W).
(4) Take two adjacent vertices, vi,r,vi,s∈V((Cm)i)−{vi,m−12,vi,m+12},r,s∈[1,m],r≠s. The resolver vk,m∈W, k∈[2,n] and k≠i, we know that d(vi,r,vi,1)≠d(vi,s,vi,1) such that
d(vi,r,vk,m)=d(vi,r,vi,1)+d(vi,1,vk,m), | (45) |
d(vi,s,vk,m)=d(vi,s,vi,1)+d(vi,1,vk,m). | (46) |
Therefore, d(vi,r,vk,m)≠d(vi,s,vk,m) so that rm(vi,r|W)≠rm(vi,s|W).
(5) The terminal vertex indices cycle graph Cn,n is odd such that the distance between two adjacent terminal vertex to the resolver vertex is distinct, d(vk,1,vi,m)≠d(vl,1,vi,m) for i∉{1,3,4} and d(vk,1,vi,m−1)≠d(vl,1,vi,m−1) for k,l∈[2,n] and i∈{1,3,4} so that rm(vk,1|W)≠rm(vl,1|W).
(6) The vertex vi,1 adjacent to vi,2, d(vi,1,vi,m)≠d(vi,2,vi,m) and
d(vi,1,vk,m)=d(vi,1,vi,1)+d(vi,1,vk,m)=d(vi,1,vk,m), | (47) |
d(vi,2,vk,m)=d(vi,2,vi,1)+d(vi,1,vk,m)=1+d(vi,1,vk,m),k≠i. | (48) |
It is clear that d(vi,1,vk,m)≠d(vi,2,vk,m), we know that rm(vi,1|W)={d(vi,1,vi,m),d(vi,1,vk,m)}≠{d(vi,2,vi,m),d(vi,2,vk,m)=rm(vi,2|W)}.
Hence, rm(vi,j|W)≠rm(vk,l|W) for vi,j adjacent to vk,l for i∈[1,n]. Consequently, W is a local m-resolving set of Cn⊳Cm.
Furthermore, we will prove that W is the local m-resolving set with minimum cardinality. Take any set S⊂V(Cn⊳Cm) with |S|<|W|. Let |S|=n−1, there are at least one cycle leaves which haven't resolver. For m odd, The number of leave vertex in every cycle leaves is even, such that the two adjacent vertices vk,m+12,vk,m+32 in (Cm)k,k∈[1,m] have the same distance to the terminal vertex, d(vk,m+12,vk,1)=m+12−1=m−12 and d(vk,m+32,vk,m)=m+1−m+32=m−12. We know that
d(vk,m+12,vi,m)=d(vk,m+12,vk,1)+d(vk,1,vi,m)=m−12+d(vk,1,vi,m), | (49) |
d(vk,m+32,vi,m)=d(vk,m+32,vk,1)+d(vk,1,vi,m)=m−12+d(vk,1,vi,m). | (50) |
It is clear that d(vk,m+12,vi,m)=d(vk,m+32,vi,m) such that rm(vk,m+12|W)=rm(vk,m+32|W). Therefore, S is not a local m-resolving set of Cn⊳Cm. Thus, mdl(Cn⊳Cm)=n. This completes the proof.
We have found the sharpest bounds of LMD of the tree comb cycle, and we get the exact value of the cycle comb tree, tree comb tree, and cycle comb cycle. There are some open problems with this research as follows.
Open Problem 1. Determine the bounds of local multiset dimension of G⊳H for G and H are any graphs.
Open Problem 2. Determine the local multiset dimension of G with mdl(G)=n−1,n−2.
Open Problem 3. Characterized mdl(G)−ldim(G)=0.
We gratefully acknowledgment the support from Universitas Jember and Universitas Airlangga for the year 2023.
There is no conflict of interest in this research.
[1] | S. Khuller, B. Raghavachari, A. Rosenfeld, Localization in graphs, 1994. Available from: http://hdl.handle.net/1903/655. |
[2] |
G. Chartrand, L. Eroh, M. A. Johnson, O. R. Oellermann, Resolvability in graphs and the metric dimension of a graph, Discrete Appl. Math., 105 (2000), 99–113. https://doi.org/10.1016/S0166-218X(00)00198-0 doi: 10.1016/S0166-218X(00)00198-0
![]() |
[3] | F. Okamoto, B. Phinezy, P. Zhang, The Local metric dimension of a graph, Math. Bohem., 135 (2010), 239–255. |
[4] | R. Simanjuntak, P. Siagian, T. Vetrik, The multiset dimension of graphs, 2017. Available from: https://doi.org/10.48550/arXiv.1711.00225. |
[5] | R. Alfarisi, Dafik, A. I. Kristiana, I. H. Agustin, The local multiset dimension of graphs, IJET 8 (2019), 120–124. |
[6] |
R. Alfarisi, Y. Lin, J. Ryan, Dafik, I. H. Agustin, A note on multiset dimension and local multiset dimension of graphs, Stat., Optim. & Inf. Comput., 8 (2020), 890–901. https://doi.org/10.19139/soic-2310-5070-727 doi: 10.19139/soic-2310-5070-727
![]() |
[7] |
R. Adawiyah, Dafik, I. H. Agustin, R. M. Prihandini, R. Alfarisi, E. R. Albirri, On the local multiset dimension of an m-shadow graph, J. Phys.: Conf. Ser., 1211 (2019), 012006. https://doi.org/10.1088/1742-6596/1211/1/012006 doi: 10.1088/1742-6596/1211/1/012006
![]() |
[8] |
R. Alfarisi, M. I. Utoyo, Dafik, Local multiset dimension of related cycle graphs, AIP Conf. Proc., 2391 (2022), 080008. https://doi.org/10.1063/5.0072516 doi: 10.1063/5.0072516
![]() |
[9] |
R. Adawiyah, R. M. Prihandini, E. R. Albirri, Dafik, I. H. Agustin, R. Alfarisi, The local multiset dimension of a unicyclic graph, IOP Conf. Ser.: Earth Environ. Sci., 243 (2019), 012075. https://doi.org/10.1088/1755-1315/243/1/012075 doi: 10.1088/1755-1315/243/1/012075
![]() |
[10] | H. Iswadi, E. T. Baskoro, A. N. M. Salman, R. Simanjuntak, The resolving graph of amalgamation of cycles, Utilitas Math., 83 (2010), 121–132. |
[11] | R. Diestel, Graph theory, Heidelberg: Springer, 2016. |
[12] | R. Alfarisi, L. Susilowati, Dafik, The Local multiset resolving of graphs, 2022, In press. |
[13] | S. W. Saputro, N. Mardiana, I. A. Purwasih, The metric dimension of comb product graph, Mat. Vestn., 4 (2017), 248–258. |
1. | Ridho Alfarisi, Liliek Susilowati, Arika Indah Kristiana, Local multiset dimension of corona product on tree graphs, 2024, 16, 1793-8309, 10.1142/S1793830923500921 |
Vertices v | Representation rm(v|W) | Conditions |
v1,j | {j−1,j+1} | j∈[1,m2+1] |
v1,j | {m−j+1,m−j+3} | j∈[m2+2,m] |
v2,j | {j,j} | j∈[1,m2] |
v2,j | {m−j,m−j+2} | j∈[m2+1,m] |
vi,j | {i+j−2,i+j−2} | i∈[3,n+12],j∈[1,m2+1] |
vi,j | {m−j+i,m−j+i} | i∈[3,n+12],j∈[m2+2,m] |
vi,j | {n−i+j,i+j−2} | i=n+32,j∈[1,m2+1] |
vi,j | {m+n−j−i+2,m−j+i} | i=n+32,j∈[m2+2,m] |
vi,j | {n−i+j,n−i+j+2} | i∈[n+52,n],j∈[1,m2+1] |
vi,j | {m+n−j−i+2,m+n−j−i+4} | i∈[n+52,n],j∈[m2+2,m] |
Vertices v | Representation rm(v|W) | Conditions |
v1,j | {j−1,j+1} | j∈[1,m2+1] |
v1,j | {m−j+1,m−j+3} | j∈[m2+2,m] |
v2,j | {j,j} | j∈[1,m2] |
v2,j | {m−j,m−j+2} | j∈[m2+1,m] |
vi,j | {i+j−2,i+j−2} | i∈[3,n+12],j∈[1,m2+1] |
vi,j | {m−j+i,m−j+i} | i∈[3,n+12],j∈[m2+2,m] |
vi,j | {n−i+j,i+j−2} | i=n+32,j∈[1,m2+1] |
vi,j | {m+n−j−i+2,m−j+i} | i=n+32,j∈[m2+2,m] |
vi,j | {n−i+j,n−i+j+2} | i∈[n+52,n],j∈[1,m2+1] |
vi,j | {m+n−j−i+2,m+n−j−i+4} | i∈[n+52,n],j∈[m2+2,m] |