Developing model checkers using PAT

Abstract

During the last two decades, model checking has emerged as an effective system analysis technique complementary to simulation and testing. Many model checking algorithms and state space reduction techniques have been proposed. Although it is desirable to have dedicated model checkers for every language (or application domain), implementing one with effective reduction techniques is rather challenging. In this work, we present a generic and extensible framework PAT, which facilitates users to build customized model checkers. PAT provides a library of state-of-art model checking algorithms as well as support for customizing language syntax, semantics, state space reduction techniques, graphic user interfaces, and even domain specific abstraction techniques. Based on this design, model checkers for concurrent systems, real-time systems, probabilistic systems and Web Services are developed inside the PAT framework, which demonstrates the practicality and scalability of our approach. © 2010 Springer-Verlag Berlin Heidelberg.