An Efficient Approach to Solving Random k-sat Problems

Proving that a propositional formula is contradictory or unsatisfiable is a fundamental task in automated reasoning. This task is coNP-complete. Efficient algorithms are therefore needed when formulae are hard to solve. Random $k-$sat formulae provide a test-bed for algorithms because experiments that have become widely popular show clearly that these formulae are consistently difficult for any known algorithm. Moreover, the experiments show a critical value of the ratio of the number of clauses to the number of variables around which the formulae are the hardest on average. This critical value also corresponds to a ‘phase transition’ from solvability to unsolvability. The question of whether the formulae located around or above this critical value can efficiently be proved unsatisfiable on average (or even for a.e. formula) remains up to now one of the most challenging questions bearing on the design of new and more efficient algorithms. New insights into this question could indirectly benefit the solving of formulae coming from real-world problems, through a better understanding of some of the causes of problem hardness. In this paper we present a solving heuristic that we have developed, devoted essentially to proving the unsatisfiability of random $k-$sat formulae and inspired by recent work in statistical physics. Results of experiments with this heuristic and its evaluation in two recent sat competitions have shown a substantial jump in the efficiency of solving hard, unsatisfiable random $k-$sat formulae.

[1]  Olivier Bailleux,et al.  Efficient CNF Encoding of Boolean Cardinality Constraints , 2003, CP.

[2]  Thomas Stützle,et al.  Local Search Algorithms for SAT: An Empirical Evaluation , 2000, Journal of Automated Reasoning.

[3]  Laurent Simon,et al.  Fifty-Five Solvers in Vancouver: The SAT 2004 Competition , 2004, SAT (Selected Papers.

[4]  Donald W. Loveland,et al.  A machine program for theorem-proving , 2011, CACM.

[5]  Gilles Dequen,et al.  The Non-existence of (3, 1, 2)-Conjugate Orthogonal Idempotent Latin Square of Order 10 , 2001, CP.

[6]  John Franco,et al.  The Probability of Pure Literals , 1999, J. Log. Comput..

[7]  Lakhdar Sais,et al.  Automatic Extraction of Functional Dependencies , 2004, SAT.

[8]  Michael E. Saks,et al.  The Efficiency of Resolution and Davis--Putnam Procedures , 2002, SIAM J. Comput..

[9]  Simona Cocco,et al.  Heuristic average-case analysis of the backtrack resolution of random 3-satisfiability instances , 2004, Theor. Comput. Sci..

[10]  James M. Crawford,et al.  Experimental Results on the Crossover Point inSatis ability , 1993 .

[11]  Uriel Feige,et al.  Resolution lower bounds for the weak pigeon hole principle , 2002, Proceedings 17th IEEE Annual Conference on Computational Complexity.

[12]  Jon W. Freeman.  Hard Random 3-SAT Problems and the Davis-Putnam Procedure , 1996, Artif. Intell..

[13]  Sharad Malik,et al.  Chaff: engineering an efficient SAT solver , 2001, Proceedings of the 38th Design Automation Conference (IEEE Cat. No.01CH37232).

[14]  Chu Min Li,et al.  Heuristics Based on Unit Propagation for Satisfiability Problems , 1997, IJCAI.

[15]  Olivier Dubois,et al.  The 3-XORSAT threshold , 2002, The 43rd Annual IEEE Symposium on Foundations of Computer Science, 2002. Proceedings..

[16]  Bart Selman,et al.  Ten Challenges in Propositional Reasoning and Search , 1997, IJCAI.

[17]  Alan Smaill,et al.  Backbone Fragility and the Local Search Cost Peak , 2000, J. Artif. Intell. Res..

[18]  Tomasz Jurdzinski,et al.  Some Results on Random Unsatisfiable k-Sat Instances and Approximation Algorithms Applied to Random Structures , 2002, Combinatorics, Probability and Computing.

[19]  Olivier Dubois,et al.  Length of Prime Implicants and Number of Solutions of Random CNF Formulae , 1999, Theor. Comput. Sci..

[20]  Gilles Dequen,et al.  A backbone-search heuristic for efficient solving of hard 3-SAT formulae , 2001, IJCAI.

[21]  Holger H. Hoos,et al.  An adaptive noise mechanism for walkSAT , 2002, AAAI/IAAI.

[22]  J. Freeman.  Improvements to propositional satisfiability search algorithms , 1995 .

[23]  M. Mézard,et al.  Analytic and Algorithmic Solution of Random Satisfiability Problems , 2002, Science.

[24]  Gilles Dequen,et al.  kcnfs: An Efficient Solver for Random k-SAT Formulae , 2003, SAT.

[25]  Sharad Malik,et al.  Efficient conflict driven learning in a Boolean satisfiability solver , 2001, IEEE/ACM International Conference on Computer Aided Design. ICCAD 2001. IEEE/ACM Digest of Technical Papers (Cat. No.01CH37281).

[26]  Riccardo Zecchina,et al.  Survey propagation: An algorithm for satisfiability , 2002, Random Struct. Algorithms.

[27]  Yacine Boufkhad,et al.  A General Upper Bound for the Satisfiability Threshold of Random r-SAT Formulae , 1997, J. Algorithms.

[28]  Bart Selman,et al.  Ten Challenges Redux: Recent Progress in Propositional Reasoning and Search , 2003, CP.

[29]  Jacques Carlier,et al.  SAT versus UNSAT , 1993, Cliques, Coloring, and Satisfiability.

[30]  Toby Walsh,et al.  Backbones in Optimization and Approximation , 2001, IJCAI.

[31]  Laurent Simon,et al.  The Essentials of the SAT 2003 Competition , 2003, SAT.

[32]  Chu Min Li,et al.  Look-Ahead Versus Look-Back for Satisfiability Problems , 1997, CP.

[33]  Rémi Monasson,et al.  2+p-SAT: Relation of typical-case complexity to the nature of the phase transition , 1999, Random Struct. Algorithms.

[34]  Bart Selman,et al.  Noise Strategies for Improving Local Search , 1994, AAAI.