Refining Restarts Strategies for SAT and UNSAT

Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 7514)


So-called Modern SAT solvers are built upon a few – but essential – ingredients: branching, learning, restarting and clause database cleaning. Most of them have been greatly improved since their first introduction, more than ten years ago. In many cases, the initial reasons that lead to their introduction do not explain anymore their current usage (for instance: very rapid restarts, aggressive clause database cleaning). Modern SAT solvers themselves share fewer and fewer properties with their ancestor, the classical backtrack search DPLL procedure.

In this paper, we explore restart strategies in the light of a new vision of SAT solvers. Following the successful results of Glucose, we consider CDCL solvers as resolution-based producers of clauses. We show that this vision is particularly salient for targeting UNSAT formulae. In a second part, we show how detecting sudden increases in the number of variable assignments can help the solver to target SAT instances too. By varying our restart strategy, we show an important improvement over Glucose 2.0, the winner of the 2011 SAT Competition, category Application SAT+UNSAT formulae. Finally we would like to point out that this new version of Glucose was the winner of the SAT Challenge 2012.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Audemard, G., Simon, L.: Predicting learnt clauses quality in modern SAT solvers. In: Proceedings of IJCAI, pp. 399–404 (2009)Google Scholar
  2. 2.
    Biere, A.: Adaptive Restart Strategies for Conflict Driven SAT Solvers. In: Kleine Büning, H., Zhao, X. (eds.) SAT 2008. LNCS, vol. 4996, pp. 28–33. Springer, Heidelberg (2008)CrossRefGoogle Scholar
  3. 3.
    Biere, A.: Picosat essentials. JSAT 4, 75–97 (2008)zbMATHGoogle Scholar
  4. 4.
    Davis, M., Logemann, G., Loveland, D.: A machine program for theorem-proving. Communication of ACM 5(7), 394–397 (1962)MathSciNetCrossRefzbMATHGoogle Scholar
  5. 5.
    Eén, N., Sörensson, N.: An Extensible SAT-solver. In: Giunchiglia, E., Tacchella, A. (eds.) SAT 2003. LNCS, vol. 2919, pp. 502–518. Springer, Heidelberg (2004)CrossRefGoogle Scholar
  6. 6.
    Gomes, C., Selman, B., Crato, N., Kautz, H.: Heavy-tailed phenomena in satisfiability and constraint satisfaction problems. Journal of Automated Reasoning 24, 67–100 (2000)MathSciNetCrossRefzbMATHGoogle Scholar
  7. 7.
    Huang, J.: The effect of restarts on the efficiency of clause learning. In: Proceedings of IJCAI, pp. 2318–2323 (2007)Google Scholar
  8. 8.
    Le Berre, D., Jarvisalo, M., Roussel, O., Simon, L.: SAT competitions (2002-2011),
  9. 9.
    Luby, M., Sinclair, A., Zuckerman, D.: Optimal speedup of las vegas algorithms. In: Israel Symposium on Theory of Computing Systems, pp. 128–133 (1993)Google Scholar
  10. 10.
    Moskewicz, M., Madigan, C., Zhao, Y., Zhang, L., Malik, S.: Chaff: Engineering an efficient SAT solver. In: Proceedings of DAC, pp. 530–535 (2001)Google Scholar
  11. 11.
    Nabeshima, H., Iwanuma, K., Inoue, K.: Glueminisat2.2.5. System Desciption,
  12. 12.
    Pipatsrisawat, K., Darwiche, A.: A Lightweight Component Caching Scheme for Satisfiability Solvers. In: Marques-Silva, J., Sakallah, K.A. (eds.) SAT 2007. LNCS, vol. 4501, pp. 294–299. Springer, Heidelberg (2007)CrossRefGoogle Scholar
  13. 13.
    Silva, J.M., Sakallah, K.: Grasp - a new search algorithm for satisfiability. In: ICCAD, pp. 220–227 (1996)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2012

Authors and Affiliations

  1. 1.CRIL / CNRS UMR8188Univ Lille-Nord de FranceLensFrance
  2. 2.LRI / CNRS UMR8623Univ Paris-SudOrsayFrance

Personalised recommendations