Generalization Strategies for the Verification of Infinite State Systems

We present a comparative evaluation of some generalization strategies which are applied by a method for the automated verification of infinite state reactive systems. The verification method is based on (1) the specialization of the constraint logic program which encodes the system with respect to the initial state and the property to be verified, and (2) a bottom-up evaluation of the specialized program. The generalization strategies are used during the program specialization phase for controlling when and how to perform generalization. Selecting a good generalization strategy is not a trivial task because it must guarantee the termination of the specialization phase itself, and it should be a good balance between precision and performance. Indeed, a coarse generalization strategy may prevent one to prove the properties of interest, while an unnecessarily precise strategy may lead to high verification times. We perform an experimental evaluation of various generalization strategies on several infinite state systems and properties to be verified.