• DocumentCode
    630394
  • Title

    Searching for Monochromatic-Square-Free Ramsey Grid Colorings via SAT Solvers

  • Author

    Walton, Paul ; Wing Ning Li

  • Author_Institution
    Comput. Sci. & Comput. Eng., Univ. of Arkansas, Fayetteville, AR, USA
  • fYear
    2013
  • fDate
    24-26 June 2013
  • Firstpage
    1
  • Lastpage
    5
  • Abstract
    Monochromatic-square-free Ramsey grid coloring is a challenging problem to solve computationally. It relates to Ramsey-theoretic combinatorics and multiple party communication protocol complexity. Recently, using a parallel search algorithm on cluster based super computers, a 2-color 14×14 solution was found in 2.5 hours. In this paper, we report on another approach to solve the monochromatic-square-free Ramsey grid coloring problem. The approach first reduces monochromatic-square-free Ramsey grid coloring to satisfiability, and then applies an existing SAT solver to solve it. Using a SAT solver on a sequential machine, we found another 2-color 14×14 solution in about two seconds, which is more than four thousand times faster that the parallel algorithm using 288 cores. We also showed that no 2-color 15×15 solution exists, which answered the open question whether a 15×15 grid graph was monochromatic-square-free 2-colorable or not.
  • Keywords
    communication complexity; computability; graph colouring; parallel algorithms; parallel machines; protocols; search problems; 2-color 14x14 solution; Ramsey-theoretic combinatorics; SAT solvers; cluster based super computers; grid graph; monochromatic-square-free Ramsey grid coloring problem; multiple party communication protocol complexity; parallel algorithm; parallel search algorithm; satisfiability; Color; Complexity theory; Computer science; Computers; Educational institutions; Parallel algorithms; Standards;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Information Science and Applications (ICISA), 2013 International Conference on
  • Conference_Location
    Suwon
  • Print_ISBN
    978-1-4799-0602-4
  • Type

    conf

  • DOI
    10.1109/ICISA.2013.6579363
  • Filename
    6579363