CELL: A compositional verification framework

Abstract

This paper presents CELL, a comprehensive and extensible framework for compositional verification of concurrent and real-time systems based on commonly used semantic models. For each semantic model, CELL offers three libraries, i.e., compositional verification paradigms, learning algorithms and model checking methods to support various state-of-the-art compositional verification approaches. With well-defined APIs, the framework could be applied to build customized model checkers. In addition, each library could be used independently for verification and program analysis purposes. We have built three model checkers with CELL. The experimental results show that the performance of these model checkers can offer similar or often better performance compared to the state-of-the-art verification tools. © 2013 Springer International Publishing.