Strengthening properties using abstraction refinement

Model checking is an automated formal method for verifying whether a finite-state system satisfies a user-supplied specification. The usefulness of the verification result depends on how well the specification distinguishes intended from non-intended system behavior. Vacuity is a notion that helps formalize this distinction in order to improve the user´s understanding of why a property is satisfied. The goal of this paper is to expose vacuity in a property in a way that increases our knowledge of the design. Our approach, based on abstraction refinement, computes a maximal set of atomic subformula occurrences that can be strengthened without compromising satisfaction. The result is a shorter and stronger and thus, generally, more valuable property. We quantify the benefits of our technique on a substantial set of circuit benchmarks.