Compositional Verification of an Object-Based Model for Reactive Systems

Reactive systems have an ongoing interaction with their environment, accepting requests and producing responses. Correct and reliable construction of reactive systems is particularly important and challenging. So, using a formal method to establish mathematical proof of correctness of such systems is necessary. A formal method consists of three major components: a model for describing the behavior of the system, a specification language to embody correctness requirements, and an analysis method to verify the behavior against the correctness requirements [8]. We choose an object-based model as the modelling language to describe the system, linear temporal logic as the specification language, and compositional verification for verifying that the model meets its specification. In system verification, a main obstacle to the use of automatic methods is the state-explosion problem, which is the exponential increase in the number of system states caused by a linear increase in the number of system components or variables [6]. Compositional verification attempts to overcome the stateexplosion problem by exploiting the modular structure that is naturally present in many system designs. Compositional verification can decrease the complexity of problem when the model is naturally decomposable [10]. So, a model consisting of inherently independent modules is suitable for compositional verification. Object-based modelling is based on abstraction, encapsulation, and information hiding, using modules with high independence. Therefore, we choose an object-based model.