Roman dominating set

From HandWiki
Short description: Type of dominating set in graph theory
File:Roman domination.svg
An assignment of the weights 0, 1 or 2 to each vertex such that each vertex with weight 0 is adjacent to at least one vertex of weight 2 is called a Roman dominating function.

In graph theory, a Roman dominating set (RDS) is a special type of dominating set inspired by historical military defense strategies of the Roman Empire. The concept models a scenario where cities (vertices) can be defended by legions stationed either within the city or in neighboring cities. A city is considered secure if it either has at least one legion stationed there, or if it has no legions but is adjacent to a city that has at least two legions, allowing one legion to be sent for defense while leaving the original city still protected.

The Roman domination number of a graph measures the minimum total number of legions needed to protect all cities according to this strategy.

Definition

Let G=(V,E) be a graph. A Roman dominating function (RDF) is a function f:V→{0,1,2} such that for every vertex v with f(v)=0, there exists a vertex u adjacent to v with f(u)=2.[1]

The weight of a Roman dominating function f is w(f)=∑v∈Vf(v). The Roman domination number γR(G) is the minimum weight among all Roman dominating functions for G.

Equivalently, let (V0,V1,V2) be an ordered partition of V where Vi={v∈V:f(v)=i}. Then f is a Roman dominating function if and only if every vertex in V0 is adjacent to at least one vertex in V2.[1]

Examples

For the complete graph Kn with n≥2, γR(Kn)=2, achieved by assigning 2 to any single vertex and 0 to all others.

For the path graph Pn and cycle graph Cn, γR(Pn)=γR(Cn)=⌈2n/3⌉.[1]

For the empty graph K‾n, γR(K‾n)=n, since each vertex must be assigned at least 1.

For the complete n-partite graph Km1,m2,…,mn with partition sizes m1≤m2≤…≤mn:[1]

  • γR(Km1,…,mn)=2 if m1=1.
  • γR(Km1,…,mn)=3 if m1=2.
  • γR(Km1,…,mn)=4 if m1≥3.

Basic properties

Several properties of Roman domination were established by Cockayne et al.:[1]

  • For any graph G, γ(G)≤γR(G)≤2γ(G), where γ(G) is the domination number.
  • γ(G)=γR(G) if and only if G is the empty graph.
  • If G has a vertex of degree n−1, then γR(G)=2.
  • For any Roman dominating function f=(V0,V1,V2):
    • The subgraph induced by V1 has maximum degree at most 1.
    • No edge joins V1 and V2.
    • Each vertex in V0 is adjacent to at most two vertices in V1.
    • V2 is a dominating set for the subgraph induced by V0∪V2.

A graph G is called a Roman graph if γR(G)=2γ(G).[2] This occurs if and only if G has a Roman dominating function of minimum weight with V1=∅.

Roman domination value

The Roman domination value of a vertex extends the concept of Roman domination by considering how many minimum Roman dominating functions assign positive values to that vertex.[3]

For a graph G, let F be the set of all γR(G)-functions (Roman dominating functions of minimum weight). For a vertex v∈V, the Roman domination value RG(v) is defined as:

RG(v)=∑f∈Ff(v)

Some basic properties of Roman domination value are known:[3]

  • 0≤RG(v)≤2τR(G), where τR(G) is the number of γR(G)-functions
  • ∑v∈V(G)RG(v)=τR(G)γR(G)
  • If there is a graph isomorphism mapping vertex v in G to vertex v′ in G′, then RG(v)=RG′(v′)

Extremal problems

Several extremal results have been established for Roman domination numbers.

For any connected n-vertex graph G with n≥3, γR(G)≤4n/5.[4] Equality holds if and only if G is C5 or obtained from n/5 copies of P5 by adding a connected subgraph on the set of centers.

For any n-vertex graph G with n≥3, 5≤γR(G)+γR(G‾)≤n+3.[4]

For any n-vertex graph G with n≥160, γR(G)γR(G‾)≤16n/5.[4]

If G is a connected n-vertex graph with δ(G)≥2 and n≥9, then γR(G)≤8n/11.[4]

Algorithms and complexity

The decision problem for Roman domination is NP-complete, even when restricted to bipartite, chordal, or planar graphs.[1] However, polynomial-time algorithms exist for computing the Roman domination number on interval graphs, cographs, and strongly chordal graphs.[2]

See also

References

  1. ↑ 1.0 1.1 1.2 1.3 1.4 1.5 Cockayne, E. J.; Dreyer, P. A.; Hedetniemi, S. M.; Hedetniemi, S. T. (2004), "Roman domination in graphs", Discrete Mathematics 278 (1–3): 11–22, doi:10.1016/j.disc.2003.06.004 
  2. ↑ 2.0 2.1 Fu, Xueliang; Yang, Yuansheng; Jiang, Baoqi (2009), "Roman domination in regular graphs", Discrete Mathematics 309 (6): 1528–1537, doi:10.1016/j.disc.2008.03.006 
  3. ↑ 3.0 3.1 Pushpam, P. R. L.; Sampath, P. (2024), "Roman domination value in graphs", Communications in Combinatorics and Optimization, doi:10.22049/cco.2024.28899.1769, https://comb-opt.azaruniv.ac.ir/article_14880_06114a6d64b7ed2996ca263b2fc2463f.pdf 
  4. ↑ 4.0 4.1 4.2 4.3 Chambers, E. W.; Kinnersley, W.; Prince, N.; West, D. B. (2009), "Extremal problems for Roman domination", SIAM Journal on Discrete Mathematics 23 (3): 1575–1586, doi:10.1137/070699688, https://digitalcommons.imsa.edu/cgi/viewcontent.cgi?article=1002&context=math_pr