Solving SAT for CNF Formulas with a One-Sided Restriction on Variable Occurrences

  • Daniel Johannsen
  • Igor Razgon
  • Magnus Wahlström
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5584)


In this paper we consider the class of boolean formulas in Conjunctive Normal Form (CNF) where for each variable all but at most d occurrences are either positive or negative. This class is a generalization of the class of CNF formulas with at most d occurrences (positive and negative) of each variable which was studied in [Wahlström, 2005].

Applying complement search [Purdom, 1984], we show that for every d there exists a constant \(\gamma_d<2-\frac{1}{2d+1}\) such that satisfiability of a CNF formula on n variables can be checked in runtime \({\ensuremath{{O}}}(\gamma_d^n)\) if all but at most d occurrences of each variable are either positive or negative. We thoroughly analyze the proposed branching strategy and determine the asymptotic growth constant γ d more precisely. Finally, we show that the trivial \({\ensuremath{{O}}}(2^n)\) barrier of satisfiability checking can be broken even for a more general class of formulas, namely formulas where the positive or negative literals of every variable have what we will call a d–covering.

To the best of our knowledge, for the considered classes of formulas there are no previous non-trivial upper bounds on the complexity of satisfiability checking.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Alon, N., Moshkovitz, D., Safra, S.: Algorithmic construction of sets for k-restrictions. ACM Trans. Algorithms 2(2), 153–177 (2006)MathSciNetCrossRefMATHGoogle Scholar
  2. 2.
    Calabro, C., Impagliazzo, R., Paturi, R.: A duality between clause width and clause density for SAT. In: Annual IEEE Conference on Computational Complexity, pp. 252–260 (2006)Google Scholar
  3. 3.
    Dantsin, E., Goerdt, A., Hirsch, E.A., Kannan, R., Kleinberg, J.M., Papadimitriou, C.H., Raghavan, P., Schöning, U.: A deterministic (2 − 2/(k + 1))n algorithm for k-SAT based on local search. Theoretical Computer Science 289(1), 69–83 (2002)MathSciNetCrossRefMATHGoogle Scholar
  4. 4.
    Paturi, R., Pudlák, P., Saks, M.E., Zane, F.: An improved exponential-time algorithm for k-sat. J. ACM 52(3), 337–364 (2005)MathSciNetCrossRefMATHGoogle Scholar
  5. 5.
    Purdom, P.W.: Solving satisfiability with less searching. IEEE Trans. Pattern Anal. Machine Intell. 6(4), 510–513 (1984)CrossRefMATHGoogle Scholar
  6. 6.
    Wahlström, M.: An algorithm for the SAT problem for formulae of linear length. In: Brodal, G.S., Leonardi, S. (eds.) ESA 2005. LNCS, vol. 3669, pp. 107–118. Springer, Heidelberg (2005)CrossRefGoogle Scholar
  7. 7.
    Wahlström, M.: Faster exact solving of SAT formulae with a low number of occurrences per variable. In: Bacchus, F., Walsh, T. (eds.) SAT 2005. LNCS, vol. 3569, pp. 309–323. Springer, Heidelberg (2005)CrossRefGoogle Scholar
  8. 8.
    Wolfram, D.: Solving generalized Fibonacci recurrences. The Fibonacci Quarterly 36(2), 129–145 (1998)MathSciNetMATHGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2009

Authors and Affiliations

  • Daniel Johannsen
    • 1
  • Igor Razgon
    • 2
  • Magnus Wahlström
    • 1
  1. 1.Max-Planck-Institut für InformatikSaarbrückenGermany
  2. 2.Cork Constraint Computation CentreUniversity College CorkIreland

Personalised recommendations