the all guards move eternal domination number for 3 n
play

The (all guards move) Eternal Domination number for 3 n Grids - PowerPoint PPT Presentation

The (all guards move) Eternal Domination number for 3 n Grids Margaret-Ellen Messinger Mount Allison University New Brunswick, CANADA with A.Z. Delaney (Mt.A.), S. Finbow (St.F.X.), M. van Bommel (St.F.X.) Margaret-Ellen Messinger (MtA)


  1. The (all guards move) Eternal Domination number for 3 ⇥ n Grids Margaret-Ellen Messinger Mount Allison University New Brunswick, CANADA with A.Z. Delaney (Mt.A.), S. Finbow (St.F.X.), M. van Bommel (St.F.X.) Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 1 / 27

  2. The Eternal Dominating Set Problem • Deployed 4 powerful field armies (each comprised of 6 legions) over 8 regions • An FA was considered capable of deploying to protect an adjacent region only if it moved from a region where there was at least one other FA to help launch it. [ReVelle & Rosing] • Consider a region to be secure if it has an FA stationed at it and securable if an FA can reach it in one step. • Constantine’s strategy is known in domination theory as Roman domination . Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 2 / 27

  3. The Eternal Dominating Set Problem guards initially form a dominating set on G at each step, a vertex is attacked in a “move” for the guards, each guard may remain where it is or move to a neighbouring vertex Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 3 / 27

  4. The Eternal Dominating Set Problem guards initially form a dominating set on G at each step, a vertex is attacked in a “move” for the guards, each guard may remain where it is or move to a neighbouring vertex if the guards “move” so that a guard is located at the attacked vertex and the set of guards again forms a dominating set, then the guards have defended against the attack We wish to find the minimum number of guards to defend against � ∞ all ( G ) any possible sequence of attacks on G . Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 4 / 27

  5. The Eternal Dominating Set Problem • special case of the (cops-first) GUARDING PROBLEM • given a board [ G ; R , C ], compute the minimum number of cops that can guard the cop-region C . C ( V ( G ) and R = V ( G ) \ C ; the cops move first and are only allowed to move within the cop-region C . If the cop-region of H is V ( G ) then G has an eternal dominating set of size k if and only if k cops can guard V ( G ). ) PSPACE-hard [Fomin, Golovach, Lokshtanov 2009] Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 5 / 27

  6. • � ∞ all known for some small classes of graphs and trees • � ( G )  � ∞ all ( G )  ↵ ( G ) Open Problem Determine the classes of graphs G with � ∞ all ( G ) = � ( G ). • If G has n vertices, � ∞ all ( G ) + � ∞ all ( G )  n + 1 l m | V ( G ) | • If G connected, � ∞ all ( G )  2 all ( G )  2 � ( G ) [sharp for all values of � ] � ∞ all ( G )  2 ⌧ ( G ) [vertex cover number] � ∞ � ( G ) � 2, � ∞ all ( G )  ⌧ ( G ) � ( G ) � 2, G girth 7 or � 9, � ∞ all ( G )  ⌧ ( G ) � 1 [survey by Mynhardt, Klostermeyer] Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 6 / 27

  7. � ( P 3 ⇤ P 3 ) = 3 = � ∞ all ( P 3 ⇤ P 3 ) � ( P 3 ⇤ P 5 ) = 4 < 5 = � ∞ all ( P 3 ⇤ P 5 ) Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 7 / 27

  8. � ( P 3 ⇤ P 3 ) = 3 = � ∞ all ( P 3 ⇤ P 3 ) � ( P 3 ⇤ P 5 ) = 4 < 5 = � ∞ all ( P 3 ⇤ P 5 ) Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 8 / 27

  9. � ( P 3 ⇤ P 3 ) = 3 = � ∞ all ( P 3 ⇤ P 3 ) � ( P 3 ⇤ P 5 ) = 4 < 5 = � ∞ all ( P 3 ⇤ P 5 ) Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 9 / 27

  10. � ( P 3 ⇤ P 3 ) = 3 = � ∞ all ( P 3 ⇤ P 3 ) � ( P 3 ⇤ P 5 ) = 4 < 5 = � ∞ all ( P 3 ⇤ P 5 ) Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 10 / 27

  11. After determining that � ∞ all ( P 3 ⇤ P n ) = n for 2  n  8, Goldwasser, Klostermeyer, Mynhardt [GKM 2012] found the surprising result that � ∞ all ( P 3 ⇤ P 9 ) = 8 which yields the upper bound Theorem 8 [GKM 2012] Conjecture 2 [GKM 2012] For n � 9, For n > 9, l 4 n l 8 n m m all ( P 3 ⇤ P n )  . all ( P 3 ⇤ P n ) = 1 + . � ∞ � ∞ 9 5 Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 11 / 27

  12. Lemma 10 [Goldwasser, Klostermeyer, Mynhardt 2012] For n > 5, P 3 ⇤ P n cannot be defended if at any step, there are only four guards in the first six columns. Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 12 / 27

  13. Lemma 10 [Goldwasser, Klostermeyer, Mynhardt 2012] For n > 5, P 3 ⇤ P n cannot be defended if at any step, there are only four guards in the first six columns. Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 13 / 27

  14. Lemma 10 [Goldwasser, Klostermeyer, Mynhardt 2012] For n > 5, P 3 ⇤ P n cannot be defended if at any step, there are only four guards in the first six columns. Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 14 / 27

  15. Lemma 10 [Goldwasser, Klostermeyer, Mynhardt 2012] For n > 5, P 3 ⇤ P n cannot be defended if at any step, there are only four guards in the first six columns. Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 15 / 27

  16. Lemma 10 [Goldwasser, Klostermeyer, Mynhardt 2012] For n > 5, P 3 ⇤ P n cannot be defended if at any step, there are only four guards in the first six columns. Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 15 / 27

  17. Lemma 10 [Goldwasser, Klostermeyer, Mynhardt 2012] For n > 5, P 3 ⇤ P n cannot be defended if at any step, there are only four guards in the first six columns. Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 17 / 27

  18. Lemma 10 [Goldwasser, Klostermeyer, Mynhardt 2012] For n > 5, P 3 ⇤ P n cannot be defended if at any step, there are only four guards in the first six columns. Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 18 / 27

  19. Theorem 6 [FMvB] l m 4 n For n � 15, � ∞ all ( P 3 ⇤ P n ) � 1 + . 5 Corollary 4 [FMvB] In any eternal dominating set of P 3 ⇤ P n , for any ` � 2, the first ` columns contain at least d 4 ` − 3 5 e guards. Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 19 / 27

  20. Claim: Let E be an eternal dominating family of P 3 ⇤ P n with fewer than 1 + d 4 n 5 e guards. In every set of E , there are at least ` � 1 guards in the first ` columns, for any ` � 6. Proof: Let ` � 6 be the smallest counterexample: in every set in E , there are at least ` � 2 guards in the first ` � 1 columns, but there is a set D 2 E in which there are ` � 2 guards in the first ` columns. Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 22 / 27

  21. Claim: Let E be an eternal dominating family of P 3 ⇤ P n with fewer than 1 + d 4 n 5 e guards. In every set of E , there are at least ` � 1 guards in the first ` columns, for any ` � 6. Proof: Let ` � 6 be the smallest counterexample: in every set in E , there are at least ` � 2 guards in the first ` � 1 columns, but there is a set D 2 E in which there are ` � 2 guards in the first ` columns. • D has ` + 1 guards in the first ` + 1 columns. l 4( n � ( ` + 1)) � 3 m Using Corollary 4, | D | � ` + 1 + 5 l 4 n m By Lemma 2 [GKM], ` � 7 ) | D | � 1 + . 5 ⇤ Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 23 / 27

  22. Theorem 6 [FMvB] l m 4 n For n � 15, � ∞ all ( P 3 ⇤ P n ) � 1 + . 5 Proof: Let E be an eternal dominating family of P 3 ⇤ P n using fewer than 1 + d 4 n 5 e guards. By the Claim, for any ` � 6, there are at least ` � 1 guards in the first ` columns of every dominating set of E . This contradicts the assumption that the dominating sets of E use fewer than 1 + d 4 n 5 e guards and the result follows. ⇤ Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 24 / 27

  23. We actually do a little better: Theorems 14 and 16 [FMvB] l m l m 4 n +1 6 n +2 For n � 11, 1 +  � ∞ all ( P 3 ⇤ P n )  . 5 7 And better still: [DM 2014+] l m l m 4 n +1 4 n For n � 11, 1 +  � ∞ all ( P 3 ⇤ P n )  2 + . 5 5 Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 25 / 27

  24. Questions: • What about � ∞ all ( P n ⇤ P n ) for n � 5? • Or � ∞ all ( P m ⇤ P n ) for m , n � 5? Thanks! university Margaret-Ellen Messinger (MtA) Eternal Domination on 3 × n grids GRASTA 2014 26 / 27

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend