<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.0 20120330//EN" "JATS-archivearticle1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink">
  <front>
    <journal-meta />
    <article-meta>
      <title-group>
        <article-title>(Eternal) Vertex Cover Number of Infinite and Finite Grid Graphs (short paper)</article-title>
      </title-group>
      <contrib-group>
        <contrib contrib-type="author">
          <string-name>Tiziana Calamoneri</string-name>
          <xref ref-type="aff" rid="aff0">0</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Federico Corò</string-name>
          <xref ref-type="aff" rid="aff1">1</xref>
        </contrib>
        <aff id="aff0">
          <label>0</label>
          <institution>Sapienza University of Rome</institution>
          ,
          <country country="IT">Italy</country>
        </aff>
        <aff id="aff1">
          <label>1</label>
          <institution>University of Padova</institution>
          ,
          <country country="IT">Italy</country>
        </aff>
      </contrib-group>
      <abstract>
        <p>In the eternal vertex cover problem, mobile guards on the vertices of a graph are used to defend it against an infinite sequence of attacks on its edges by moving to neighbor vertices. The eternal vertex cover problem consists in determining the minimum number of necessary guards. Motivated by previous literature, in this paper, we study the vertex cover and eternal vertex cover problems on regular grids when passing from the infinite to the finite version of the same graphs, and we provide either coinciding or very tight lower and upper bounds on the number of necessary guards. To this aim, we generalize the notions of minimum vertex and minimum eternal vertex covers in order to be well-defined for infinite grids.</p>
      </abstract>
      <kwd-group>
        <kwd>eol&gt;Min vertex cover</kwd>
        <kwd>infinite grids</kwd>
        <kwd>Eternal Vertex Cover</kwd>
      </kwd-group>
    </article-meta>
  </front>
  <body>
    <sec id="sec-1">
      <title>1. Introduction</title>
      <p>The eternal vertex cover problem can be described in terms of a two-player game and models a
problem associated with defending the vertices of a given graph  against a sequence of attacks:
at the beginning of the game, the defender (which controls some guards lying on the vertices of
) chooses a placement of guards on a subset of vertices of , defining an initial configuration.
Subsequently, in each round of the game, the attacker attacks an edge of their choice. To repel
an attack, a guard from an incident vertex moves across the attacked edge to defend it; at the
same time, other guards may either remain where they are or move to neighbor vertices. If
the defender is able to do this, then the attack has been successfully defended, and the game
proceeds to another round, with the attacker choosing the next edge to attack. Otherwise, the
attacker wins.</p>
      <p>
        Graph protection has its historical roots in the time of the ancient Roman Empire [
        <xref ref-type="bibr" rid="ref1">1</xref>
        ]; according
to E. N. Luttwak [
        <xref ref-type="bibr" rid="ref2">2</xref>
        ], to deal with the dwindling power of the Empire, Constantine devised a
new strategy that used local troops to disrupt invasions. The idea was to deploy mobile armies
that could only protect an adjacent region if it moved from a region where there was at least
one other army to help launch it. Basically, a region is considered as safe if either has one or
more armies stationed in it, or if it can be reached in one pass. A similar strategy was then
used in World War II [
        <xref ref-type="bibr" rid="ref2">2</xref>
        ]. Currently, the problem finds applications in network security and
redundant file storage handling [
        <xref ref-type="bibr" rid="ref1">1</xref>
        ].
      </p>
      <p>If the defender is able to continue defending against any infinite sequence of attacks on 
with  guards, then we say that there is a defense strategy on  with  guards. This strategy
requires that the set of vertices containing guards is a vertex cover before and after each round.
In this case, the set of positions of all the guards in any round of the game defines a configuration,
referred to as an eternal vertex cover of  of size .</p>
      <p>The eternal vertex cover problem on a graph  requires finding the minimum value of  for
which a defense strategy on  with  guards exists. Such minimum is denoted by  ∞().</p>
      <p>Literature. The problem was first formulated in 2009 by Klostermeyer and Mynhardt [ 3] and,
in the same article, it was shown that, for any graph ,  () ≤  ∞() ≤ 2 (), where  ()
is the cardinality of a minimum vertex cover for .</p>
      <p>The problem has been deeply studied from a computational complexity point of view: deciding
whether  guards can protect all the edges of a graph is NP-hard [4] and remains NP-hard
even for internally triangulated planar graphs [5]; moreover, it is APX-hard and there is a
2-approximation algorithm; finally, exact (exponential) algorithms are given. Recently, it was
shown that the problem remains NP-hard even for bipartite graphs [6].</p>
      <p>Nevertheless, there are a number of special graphs for which the problem can be exactly
solved in polynomial time: trees [3] ( ∞( ) =  − | ( )| + 1, where ( ) is the number of
leaves of  ), cacti [7, 8], simple generalizations of trees [9] (graphs constructed by replacing
each edge of a tree with an arbitrary elementary bipartite graph or by an arbitrary clique),
perfect matching cycles on  vertices [3] ( ∞() =  () = ⌈ 2 ⌉), chordal graphs [5, 8],
chain graphs, cographs, split graphs [10], and  ×  grids [3] (whose  ∞ is 2 =  if  is
even and is ⌈ 2 ⌉ =  + 1 if ,  &gt; 1 are odd with  ≥ ).</p>
      <p>
        More in general, in [
        <xref ref-type="bibr" rid="ref1">3, 1</xref>
        ], some conditions to have  ∞ =  are provided and it seems
particularly interesting to understand in which cases these two parameters are very close. Babu
et al. [11] showed that  ∞ ≤  + 1 for locally connected graphs, which includes biconnected
chordal graphs and biconnected internally triangulated planar graphs. While providing a
polynomial time algorithm to compute  ∞ for the former, and a PTAS for the latter.
      </p>
      <p>Finally, there are a number of papers (e.g. [12, 13, 5, 14]) connecting  and  ∞ of special
graphs with other parameters such that the (eternal) domination number, the vertex connectivity
number, and the minimum cardinality of a vertex cover that contains all cut vertices.</p>
    </sec>
    <sec id="sec-2">
      <title>2. Definitions</title>
      <p>In this extended abstract we consider first infinite regular grids , i.e. infinite plane graphs in which
all faces are identical regular polygons: hexagonal, squared, and triangular grids, here denoted
by Δ, where Δ represents the degree that is equal to 3,4,6, respectively. For completeness, we
also study the regular grid of degree 8, called octagonal grid (also known as “king’s graph”),
and denoted by 8.</p>
      <p>Then we study also finite portions of infinite grid graphs, that are induced by the vertices of a
set of regular faces contained into a ℎ ×  rectangle with sides parallel to the axes and vertices at
(a) hexagonal grid
(b) squared grid
(c) triangular grid
(d) octagonal grid
integer coordinates, ℎ,  ∈ N; we will call these subgraphs as finite rectangular grids and denote
them with Δ(ℎ, ). Clearly, this concept is not well defined for octagonal grids since they
are not planar graphs, but it is not dificult to extend it (for example by momentarily removing
bottom-right to up-left edges, determining the subgraph according to the definition, and then
adding again the removed edges). Moreover, we assume the finite grids do not degenerate in
simpler structures (and hence, e.g. the cases ℎ = 1 and  = 1 are not allowed in any grid, while
the case ℎ &lt; 4 is not allowed in the hexagonal grid).</p>
      <p>We now give a formal definition of the problems we want to study on infinite and finite grids.
It is worth noting that, while the following definitions are very well known in the case of finite
graphs, we need to extend them in order to make them work on infinite grids.
Definition 1. Given a (either finite or infinite) graph  = (, ), a vertex cover for  is a set
of vertices  ⊆  that includes at least one endpoint of every edge. If  is a finite graph, the
vertex cover problem consists in finding a vertex cover of minimum cardinality, and this number
is denoted with  (). If  is an infinite grid, the vertex cover problem consists in finding a vertex
cover  such that there exists 0 &gt; 0 and, for every  ≥ 0 there is a finite rectangular grid of ,
(, ) with a vertex set  (), such that  ∩  () is a minimum vertex cover for (, ).</p>
      <p>In both cases, we will say that a solution to the vertex cover problem is a minimum vertex cover.</p>
      <p>Given a vertex cover  for , an attack may occur on a single edge  = {, }, where  ∈ ;
a defense by  to the attack on the edge is a one-to-one function  :  →  such that:
(1)  () = , and
(2) for each  ∈  ∖ {},  () ∈  [] where  [] is the close neighborhood of .</p>
      <p>Given any vertex  ∈ , we say that the guard on  shifts to  () and, by extension,  shifts
to ′ where ′ = { () s.t.  ∈ }.</p>
      <p>Since an attack of an edge whose both its extremes contain a guard can always be repelled
without changing the configuration of guards (simply swapping the position of the guards on
that edge), we only consider attacks on edges with one unguarded vertex.</p>
      <p>Definition 2. Given a (either finite or infinite) graph  = (, ), a vertex cover  for  is
an eternal vertex cover if every (possibly infinite) sequence of attacks can be defended, that is
if a defense shifts  in ′ and ′ is an eternal vertex cover. If  is a finite graph, the eternal
vertex cover problem consists in finding an eternal vertex cover of minimum cardinality, and this
number is denoted by  ∞(). If  is an infinite grid, the eternal vertex cover problem consists
in finding an eternal vertex cover  such that there exists 0 &gt; 0 and, for every  ≥ 0 there are
(possibly coinciding) finite rectangular grids of , (, ) and ′(, ) with vertex set  () and
 ′(), such that  ∩  () and ′ ∩  ′() are minimum vertex covers for (, ) and ′(, ),
respectively.</p>
      <p>In both cases, we will say that a solution to the eternal vertex cover problem is a minimum
eternal vertex cover.</p>
      <p>In order to compare the results we obtain for finite and infinite graphs, we introduce the
definitions of  and  ∞.</p>
      <p>Definition 3. Let  be a finite graph. We call  the ratio between the number of vertices in a
minimum vertex cover and the number of all the vertices of , and  ∞ the ratio between the
number of vertices in an eternal minimum vertex cover and the number of all the vertices of .</p>
      <p>Let  be an infinite grid,  and  be a minimum vertex cover and an eternal minimum vertex
cover, respectively, for . For each  &gt; 0 consider every possible finite rectangular grid (, )
and let  () be its vertex set; compute the minimum over all finite grids of the ratio between
| () ∩ | (respectively | () ∩ |) and | ()| = 2. We call  (respectively  ∞) the limit as 
goes to ∞ of this ratio.</p>
      <p>Intuitively,  and  ∞ represent the fraction of vertices that belong to a minimum cardinality
cover.</p>
    </sec>
    <sec id="sec-3">
      <title>3. Summary of the results</title>
      <p>This study arises from observing that, while the density of an eternal vertex cover of an
infinite path is roughly 1/2, we have that in the finite path  almost all vertices are needed
( ∞() =  − 1 [3]).</p>
      <p>On the other hand, it is immediate to see that any eternal vertex cover of the infinite squared
grid puts guards alternately on its vertices; this is true also in the finite case, indeed, as we have
already stated, from the literature, we know that a squared grid (i.e. the Cartesian product of
two paths) has  ∞ roughly equal to half of its number of vertices.</p>
      <p>So, not all the graphs behave in the same way when passing from the infinite to the finite
version. We wonder which is the behavior of some naturally innfiite graphs, that is whether we
get an increase of their eternal vertex cover number when we reduce to a finite portion.</p>
      <p>To this aim, we first consider infinite regular grid graphs. For each of them, we evaluate
on which portion of the vertices it is necessary to put a guard; moreover, we consider a finite
(rectangular) portion of such graphs and study the value of their  ∞. Finally, we compare our
results in infinite and finite cases. The results of this study are summarized in Tables 1 and 2.</p>
      <p>It is worth noting that, to the best of our knowledge, in the literature, no generalizations are
known for the notions of minimum vertex cover and eternal vertex cover of infinite graphs.</p>
      <p>This study concludes that having a not null second dimension helps an efective defense
strategy; indeed, the path remains the only graph for which  ∞ is completely diferent when
passing from the infinite to the finite case.
Summary of the results for finite grid graphs; the lower part of the table contains results provided in
this paper. The results of the last two rows related to  ∞ hold for ℎ ≥ .</p>
      <p>For the full version of the current paper, please refer to [15].
(2009) 235–250.
[4] F. V. Fomin, S. Gaspers, P. A. Golovach, D. Kratsch, S. Saurabh, Parameterized algorithm
for eternal vertex cover, Information Processing Letters 110 (2010) 702–706.
[5] J. Babu, V. Prabhakaran, A new lower bound for the eternal vertex cover number of graphs,</p>
      <p>J. Comb. Optim. 44 (2022) 2482–2498.
[6] N. Misra, S. Nanoti, Eternal vertex cover on bipartite and co-bipartite graphs, CoRR
abs/2201.03820 (2022). URL: https://arxiv.org/abs/2201.03820. arXiv:2201.03820.
[7] J. Babu, V. Prabhakaran, A. Sharma, A linear time algorithm for computing the eternal
vertex cover number of cactus graphs, arXiv preprint arXiv:2005.08058 (2020).
[8] J. Babu, V. Prabhakaran, A. Sharma, A substructure based lower bound for eternal vertex
cover number, Theor. Comput. Sci. 890 (2021) 87–104.
[9] H. Araki, T. Fujito, S. Inoue, On the eternal vertex cover numbers of generalized trees, IEICE
Transactions on Fundamentals of Electronics, Communications and Computer Sciences 98
[10] K. Paul, A. Pandey, Some algorithmic results for eternal vertex cover problem in graphs,
in: C. Lin, B. M. T. Lin, G. Liotta (Eds.), WALCOM: Algorithms and Computation - 17th
International Conference and Workshops, WALCOM 2023, Hsinchu, Taiwan, March 22-24,
2023, Proceedings, volume 13973 of Lecture Notes in Computer Science, Springer, 2023, pp.
242–253.
[11] J. Babu, L. S. Chandran, M. C. Francis, V. Prabhakaran, D. Rajendraprasad, J. N. Warrier,
On graphs whose eternal vertex cover number and vertex cover number coincide, Discret.</p>
      <p>Appl. Math. 319 (2022) 171–182.
[12] M. Anderson, R. Brigham, J. Carrington, R. Dutton, R. Vitray, J. Yellen, Mortal and eternal
vertex covers, Preprint (2012).
[13] M. Anderson, R. Brigham, J. Carrington, R. Dutton, R. Vitray, J. Yellen, Graphs
simultaneously achieving three vertex cover numbers, J. Combin. Math. Combin. Comput 91 (2014)
275–290.
[14] W. F. Klostermeyer, C. M. Mynhardt, Graphs with equal eternal vertex cover and eternal
domination numbers, Discrete Mathematics 311 (2011) 1371–1379.
[15] T. Calamoneri, F. Corò, (Eternal) Vertex Cover number of infinite and finite grid graphs,
arXiv preprint arXiv:2209.05102 (2022).</p>
    </sec>
  </body>
  <back>
    <ref-list>
      <ref id="ref1">
        <mixed-citation>
          [1]
          <string-name>
            <given-names>W. F.</given-names>
            <surname>Klostermeyer</surname>
          </string-name>
          ,
          <string-name>
            <given-names>C.</given-names>
            <surname>Mynhardt</surname>
          </string-name>
          ,
          <article-title>Protecting a graph with mobile guards</article-title>
          ,
          <source>Applicable Analysis and Discrete Mathematics</source>
          <volume>10</volume>
          (
          <year>2016</year>
          )
          <fpage>1</fpage>
          -
          <lpage>29</lpage>
          . strategy,
          <source>The American Mathematical Monthly</source>
          <volume>107</volume>
          (
          <year>2000</year>
          )
          <fpage>585</fpage>
          -
          <lpage>594</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref2">
        <mixed-citation>
          [2]
          <string-name>
            <given-names>C. S.</given-names>
            <surname>ReVelle</surname>
          </string-name>
          , K. E. Rosing,
          <article-title>Defendens imperium romanum: a classical problem in military</article-title>
          [3]
          <string-name>
            <given-names>W. F.</given-names>
            <surname>Klostermeyer</surname>
          </string-name>
          ,
          <string-name>
            <given-names>C. M.</given-names>
            <surname>Mynhardt</surname>
          </string-name>
          ,
          <article-title>Edge protection in graphs</article-title>
          .,
          <string-name>
            <surname>Australas</surname>
          </string-name>
          .
          <source>J Comb</source>
          .
          <volume>45</volume>
        </mixed-citation>
      </ref>
    </ref-list>
  </back>
</article>