Inês Lynce

Towards robust CNF encodings of cardinality constraints

Motivated by the performance improvements made to SAT solvers in recent years, a number of different encodings of constraints into SAT have been proposed. Concrete examples are the different SAT encodings for ≤ 1 (x1, . . . , xn) constraints. The most widely used encoding is known as the pairwise encoding, which is quadratic in the number of variables in the constraint. Alternative encodings are in general linear, and require using additional auxiliary variables. In most settings, the pairwise encoding performs acceptably well, but can require unacceptably large Boolean formulas. In contrast, linear encodings yield much smaller Boolean formulas, but in practice SAT solvers often perform unpredictably. This lack of predictability is mostly due to the large number of auxiliary variables that need to be added to the resulting Boolean formula. This paper studies one specific encoding for ≤ 1 (x1, . . . , xn) constraints, and shows how a state-of-the-art SAT solver can be adapted to overcome the problem of adding additional auxiliary variables. Moreover, the paper shows that a SAT solver may essentially ignore the existence of auxiliary variables. Experimental results indicate that the modified SAT solver becomes significantly more robust on SAT encodings involving ≤ 1 (x1, . . . , xn) constraints.

Towards efficient MUS extraction

Minimally Unsatisfiable Subformulas (MUS) find a wide range of practical applications, including product configuration, knowledge-based validation, and hardware and software design and verification. MUSes also find application in recent Maximum Satisfiability algorithms and in CNF formula redundancy removal. Besides direct applications in Propositional Logic, algorithms for MUS extraction have been applied to more expressive logics. This paper proposes two algorithms for computation of MUSes of propositional formulas in Conjunctive Normal Form (CNF). The first algorithm is optimal in its class, meaning that it requires the smallest number of calls to a SAT solver. The second algorithm extends earlier work, but implements a number of new techniques. Among these, this paper analyzes in detail the technique of recursive model rotation, which provides significant performance gains in practice. Experimental results, obtained on representative practical benchmarks, indicate that the new algorithms achieve significant performance gains with respect to state of the art MUS extraction algorithms.

A branch-and-bound algorithm for extracting smallest minimal unsatisfiable formulas

We tackle the problem of finding a smallest-cardinality MUS (SMUS) of a given formula. The SMUS provides a succinct explanation of infeasibility and is valuable for applications that rely on such explanations. We present a branch-and-bound algorithm that utilizes iterative MAXSAT solutions to generate lower and upper bounds on the size of the SMUS, and branch on specific subformulas to find it. We report experimental results on formulas from DIMACS and DaimlerChrysler product configuration suites.

On improving MUS extraction algorithms

Minimally Unsatisfiable Subformulas (MUS) find a wide range of practical applications, including product configuration, knowledge-based validation, and hardware and software design and verification. MUSes also find application in recentMaximum Satisfiability algorithms and in CNF formula redundancy removal. Besides direct applications in Propositional Logic, algorithms for MUS extraction have been applied to more expressive logics. This paper proposes two algorithms forMUS extraction. The first algorithm is optimal in its class, meaning that it requires the smallest number of calls to a SAT solver. The second algorithm extends earlier work, but implements a number of new techniques. The resulting algorithms achieve significant performance gains with respect to state of the art MUS extraction algorithms.

Probing-based preprocessing techniques for propositional satisfiability

Preprocessing is an often used approach for solving hard instances of propositional satisfiability (SAT). Preprocessing can be used for reducing the number of variables and for drastically modifying the set of clauses, either by eliminating irrelevant clauses or by inferring new clauses. Over the years, a large number of formula manipulation techniques has been proposed, that in some situations have allowed solving instances not otherwise solvable with state-of-the-art SAT solvers. This paper proposes probing-based preprocessing, an integrated approach for preprocessing propositional formulas, that for the first time integrates in a single algorithm most of the existing formula manipulation techniques. Moreover, the new unified framework can be used to develop new techniques. Preliminary experimental results illustrate that probing-based preprocessing can be effectively used as a preprocessing tool in state-of-the-art SAT solvers.

Open-WBO: A Modular MaxSAT Solver,

This paper presents open-wbo, a new MaxSAT solver. open-wbo has two main features. First, it is an open-source solver that can be easily modified and extended. Most MaxSAT solvers are not available in open-source, making it hard to extend and improve current MaxSAT algorithms. Second, open-wbo may use any MiniSAT-like solver as the underlying SAT solver. As many other MaxSAT solvers, open-wbo relies on successive calls to a SAT solver. Even though new techniques are proposed for SAT solvers every year, for many MaxSAT solvers it is hard to change the underlying SAT solver. With open-wbo, advances in SAT technology will result in a free improvement in the performance of the solver. In addition, the paper uses open-wbo to evaluate the impact of using different SAT solvers in the performance of MaxSAT algorithms.

An Overview of Backtrack Search Satisfiability Algorithms

Propositional Satisfiability (SAT) is often used as the underlying model for a significant number of applications in Artificial Intelligence as well as in other fields of Computer Science and Engineering. Algorithmic solutions for SAT include, among others, local search, backtrack search and algebraic manipulation. In recent years, several different organizations of local search and backtrack search algorithms for SAT have been proposed, in many cases allowing larger problem instances to be solved in different application domains. While local search algorithms have been shown to be particularly useful for random instances of SAT, recent backtrack search algorithms have been used for solving large instances of SAT from real-world applications. In this paper we provide an overview of backtrack search SAT algorithms. We describe and illustrate a number of techniques that have been empirically shown to be highly effective in pruning the amount of search on significant and representative classes of problem instances. In particular, we review strategies for non-chronological backtracking, procedures for clause recording and for the identification of necessary variable assignments, and mechanisms for the structural simplification of instances of SAT.

Solving Linux Upgradeability Problems Using Boolean Optimization

Managing the software complexity of package-based systems can be regarded as one of the main challenges in software architectures. Upgrades are required on a short time basis and systems are expected to be reliable and consistent after that. For each package in the system, a set of dependencies and a set of conflicts have to be taken into account. Although this problem is computationally hard to solve, efficient tools are required. In the best scenario, the solutions provided should also be optimal in order to better fulfill users requirements and expectations. This paper describes two different tools, both based on Boolean satisfiability (SAT), for solving Linux upgradeability problems. The problem instances used in the evaluation of these tools were mainly obtained from real environments, and are subject to two different lexicographic optimization criteria. The developed tools can provide optimal solutions for many of the instances, but a few challenges remain. Moreover, it is our understanding that this problem has many similarities with other configuration problems, and therefore the same techniques can be used in other domains.

Incremental Cardinality Constraints for MaxSAT

Maximum Satisfiability (MaxSAT) is an optimization variant of the Boolean Satisfiability (SAT) problem. In general, MaxSAT algorithms perform a succession of SAT solver calls to reach an optimum solution making extensive use of cardinality constraints. Many of these algorithms are non-incremental in nature, i.e. at each iteration the formula is rebuilt and no knowledge is reused from one iteration to another. In this paper, we exploit the knowledge acquired across iterations using novel schemes to use cardinality constraints in an incremental fashion. We integrate these schemes with several MaxSAT algorithms. Our experimental results show a significant performance boost for these algorithms as compared to their non-incremental counterparts. These results suggest that incremental cardinality constraints could be beneficial for other constraint solving domains.

SAT in bioinformatics: making the case with haplotype inference

Mutation in DNA is the principal cause for differences among human beings, and Single Nucleotide Polymorphisms (SNPs) are the most common mutations. Hence, a fundamental task is to complete a map of haplotypes (which identify SNPs) in the human population. Associated with this effort, a key computational problem is the inference of haplotype data from genotype data, since in practice genotype data rather than haplotype data is usually obtained. Recent work has shown that a SAT-based approach is by far the most efficient solution to the problem of haplotype inference by pure parsimony (HIPP), being several orders of magnitude faster than existing integer linear programming and branch and bound solutions. This paper proposes a number of key optimizations to the the original SAT-based model. The new version of the model can be orders of magnitude faster than the original SAT-based HIPP model, particularly on biological test data.