An Iterative Approach to Language Containment

We propose an iterative approach to formal verification by language containment. We start with some initial abstraction and then iteratively refine it, guided by the failure report from the verification tool. We show that the procedure will terminate, propose a series of heuristic aimed at reducing the size of BDD's used in the computation, and formulate several open problems that could improve efficiency of the procedure. Finally, we present and discuss some initial experimental results.