Combination model checking: approach and a case study

We present combination model checking approach using a SAT-based bounded model checker together with a BDD-based symbolic model checker to provide a more efficient counter example generation process. We provide this capability without compromising the verification capability of the symbolic model checker. The basic idea is to use the symbolic model checker to determine whether or not a property holds in the model. If the property holds, we are done. If it does not, we preempt the counterexample generation and use the SAT-based model checker for this purpose. An application of the combination approach to a version of a Flight Guidance System (FGS) from Rockwell Collins, Inc. shows huge performance gain when checking a collection of several hundred properties.