Jordi Planes

Algorithms for Weighted Boolean Optimization

The Pseudo-Boolean Optimization (PBO) and Maximum Satisfiability (MaxSAT) problems are natural optimization extensions of Boolean Satisfiability (SAT). In the recent past, different algorithms have been proposed for PBO and for MaxSAT, despite the existence of straightforward mappings from PBO to MaxSAT, and vice-versa. This papers proposes Weighted Boolean Optimization (WBO), a new unified framework that aggregates and extends PBO and MaxSAT. In addition, the paper proposes a new unsatisfiability-based algorithm for WBO, based on recent unsatisfiability-based algorithms for MaxSAT. Besides standard MaxSAT, the new algorithm can also be used to solve weighted MaxSAT and PBO, handling pseudo-Boolean constraints either natively or by translation to clausal form. Experimental results illustrate that unsatisfiability-based algorithms for MaxSAT can be orders of magnitude more efficient than existing dedicated algorithms. Finally, the paper illustrates how other algorithms for either PBO or MaxSAT can be extended to WBO.

New inference rules for Max-SAT

Exact Max-SAT solvers, compared with SAT solvers, apply little inference at each node of the proof tree. Commonly used SAT inference rules like unit propagation produce a simplified formula that preserves satisfiability but, unfortunately, solving the Max-SAT problem for the simplified formula is not equivalent to solving it for the original formula. In this paper, we define a number of original inference rules that, besides being applied efficiently, transform Max-SAT instances into equivalent Max-SAT instances which are easier to solve. The soundness of the rules, that can be seen as refinements of unit resolution adapted to Max-SAT, are proved in a novel and simple way via an integer programming transformation. With the aim of finding out how powerful the inference rules are in practice, we have developed a new Max-SAT solver, called MaxSatz, which incorporates those rules, and performed an experimental investigation. The results provide empirical evidence that MaxSatz is very competitive, at least, on random Max-2SAT, random Max-3SAT, Max-Cut, and Graph 3-coloring instances, as well as on the benchmarks from the Max-SAT Evaluation 2006.

Algorithms for maximum satisfiability using unsatisfiable cores

Many decision and optimization problems in electronic design automation (EDA) can be solved with Boolean satisfiability (SAT). Moreover, well-known extensions of SAT also find application in EDA, including pseudo-Boolean optimization, quantified Boolean formulas, multi-valued SAT and, more recently, Maximum Satisfiability (MaxSAT). Algorithms for MaxSAT are still fairly inefficient in industrial settings, in part because the most effective SAT techniques cannot be easily extended to MaxSAT. This paper proposes a novel algorithm for MaxSAT that improves existing state of the art solvers by orders of magnitude on industrial benchmarks. The new algorithm exploits modern SAT solvers, being based on the identification of unsatisfiable subformulas. Moreover, the new algorithm provides additional insights between unsatisfiable subformulas and the maximum satisfiability problem.

Iterative and core-guided MaxSAT solving: A survey and assessment

Maximum Satisfiability (MaxSAT) is an optimization version of SAT, and many real world applications can be naturally encoded as such. Solving MaxSAT is an important problem from both a theoretical and a practical point of view. In recent years, there has been considerable interest in developing efficient algorithms and several families of algorithms have been proposed. This paper overviews recent approaches to handle MaxSAT and presents a survey of MaxSAT algorithms based on iteratively calling a SAT solver which are particularly effective to solve problems arising in industrial settings. First, classic algorithms based on iteratively calling a SAT solver and updating a bound are overviewed. Such algorithms are referred to as iterative MaxSAT algorithms. Then, more sophisticated algorithms that additionally take advantage of unsatisfiable cores are described, which are referred to as core-guided MaxSAT algorithms. Core-guided MaxSAT algorithms use the information provided by unsatisfiable cores to relax clauses on demand and to create simpler constraints. Finally, a comprehensive empirical study on non-random benchmarks is conducted, including not only the surveyed algorithms, but also other state-of-the-art MaxSAT solvers. The results indicate that (i) core-guided MaxSAT algorithms in general abort in less instances than classic solvers based on iteratively calling a SAT solver and that (ii) core-guided MaxSAT algorithms are fairly competitive compared to other approaches.

Exploiting unit propagation to compute lower bounds in branch and bound Max-SAT solvers

One of the main differences between complete SAT solvers and exact Max-SAT solvers is that the former make an intensive use of unit propagation at each node of the proof tree while the latter, in order to ensure optimality, can only apply unit propagation to a restricted number of nodes. In this paper, we describe a branch and bound Max-SAT solver that applies unit propagation at each node of the proof tree to compute the lower bound instead of applying unit propagation to simplify the formula. The new lower bound captures the lower bound based on inconsistency counts that apply most of the state-of-the-art Max-SAT solvers as well as other improvements, like the start rule, that have been defined to get a lower bound of better quality. Moreover, our solver incorporates the Jeroslow-Wang variable selection heuristic, the pure literal and dominating unit clause rules, and novel preprocessing techniques. The experimental investigation we conducted to compare our solver with the most modern Max-SAT solvers provides experimental evidence that our solver is very competitive. Research partially supported by projects TIN2004-07933-C03-03 and TIC2003-00950 funded by the Ministerio de Educacion y Ciencia. The second author is supported by a grant Ramon y Cajal.

Exploiting Cycle Structures in Max-SAT

We investigate the role of cycles structures (i.e., subsets of clauses of the form

Resolution-based lower bounds in MaxSAT

\bar{l}_{1}\vee l_{2}, \bar{l}_{1}\vee l_{3},\bar{l}_{2}\vee\bar{l}_{3}

Improved exact solvers for weighted Max-SAT

) in the quality of the lower bound (LB) of modern MaxSAT solvers. Given a cycle structure, we have two options: (i) use the cycle structure just to detect inconsistent subformulas in the underestimation component, and (ii) replace the cycle structure with

A Max-SAT Solver with Lazy Data Structures

\bar{l}_{1},l_{1}\vee\bar{l}_{2}\vee\bar{l}_{3},\bar{l}_{1}\vee l_{2}\vee l_{3}

Iterative SAT Solving for Minimum Satisfiability

by applying MaxSAT resolution and, at the same time, change the behaviour of the underestimation component. We first show that it is better to apply MaxSAT resolution to cycle structures occurring in inconsistent subformulas detected using unit propagation or failed literal detection. We then propose a heuristic that guides the application of MaxSAT resolution to cycle structures during failed literal detection, and evaluate this heuristic by implementing it in MaxSatz, obtaining a new solver called MaxSatz c . Our experiments on weighted MaxSAT and Partial MaxSAT instances indicate that MaxSatz c substantially improves MaxSatz on many hard random, crafted and industrial instances.