Verification of Infinite State Systems [chapter]

Ahmed Bouajjani
2003 Lecture Notes in Computer Science  
The development of our modern societies needs more and more involvement of computers in managing highly complex and (safety-)critical tasks, e.g., in telecommunication, chemical and physical process control, transportation systems, etc. It is essential to be able to produce reliable hardware and software systems, since any erroneous behaviour can have catastrophic (economical and human) consequences. This requires rigorous methods and techniques to conceive, analyze and validate these systems.
more » ... he verification problem consists in checking whether a system satisfies its specification. During the two last decades, significant achievements have been obtained in the case of finite-state systems (systems with finitely many states). One of the main actual challenges in the domain of automated verification is the conception of methods and tools allowing to deal with verification problems beyond the finite-state framework. Such problems rise naturally as soon as we consider aspects like: • real-time constraints: timed and hybrid systems, • unbounded discrete data structures: counters, fifo-channels, stacks, etc. • parametric reasoning about families of systems, e.g., networks of processes, • process mobility, dynamic creation and destruction of processes (dynamic modification of the communication structure). In the last two years the specification and verification of infinite-state systems have attracted the attention of more and more researchers belonging to a very broad range of research communities. Both process algebras (or term rewriting systems) and automata (or finite control machines) are being used as specification formalisms. Verification problems can be reduced to checking behavioural equivalence or implementation (simulation) relations, or to checking the satisfaction of properties described in temporal logics or fixpoint calculi (model checking problems). Verification methods can be deductive (based on the use of theorem provers), or algorithmic (based on decision or semi-decision procedures). Algorithmic methods can be based on fixpoint theory, automata theory or (constrained) logic programming.
doi:10.1007/978-3-540-45220-1_7 fatcat:4j5lzdbfc5hd7ptdo4tpwvmd2m