Bounded Model Checking (BMC) is the problem of checking if a model satisfies a temporal property in paths with bounded length k. Propositional SAT-based BMC is conducted in a gradual manner, by solving a series of SAT instances corresponding to formulations of the problem with increasing k. We show how the gradual nature can be exploited for shortening the overall verification time. The concept is to reuse constraints on the search space which are deduced while checking a k instance, for speeding up the SAT checking of the consecutive k+1 instance. This technique can be seen as a generalization of 'pervasive clauses', a technique introduced by Silva and Sakallah in the context of Automatic Test Pattern Generation (ATPG). We define the general conditions for reusability of constraints, and define a simple procedure for evaluating them. This technique can theoretically be used in any solution that is based on solving a series of closely related SAT instances (instances with non-empty intersection between their set of clauses). We then continue by showing how a similar procedure can be used for restricting the search space of individual SAT instances corresponding to BMC invariant formulas. Experiments demonstrated that both techniques have consistent and significant positive effect.

Bounded Model Checking based on SAT methods has recently been introduced as a complementary technique to BDD-based Symbolic Model Checking. The basic idea is to search for a counter example in executions whose length is bounded by some integer k. The BMC problem can be efficiently reduced to a propositional satisfiability problem, and can therefore be solved by SAT methods rather than BDDs. SAT procedures are based on general-purpose heuristics that are designed for any propositional formula. We show that the unique characteristics of BMC formulas can be exploited for a variety of optimizations in the SAT checking procedure. Experiments with these optimizations on real designs proved their efficiency in many of the hard test cases, comparing to both the standard SAT procedure and a BDD-based model checker.

No abstract

No abstract