Complexity Results for Model Checking

Abstract

The complexity of model checking branching and linear time<br />temporal logics over Kripke structures has been addressed in e.g. [SC85,<br />CES86]. In terms of the size of the Kripke model and the length of the<br />formula, they show that the model checking problem is solvable in <br />polynomial time for CTL and NP-complete for L(F). The model checking<br />problem can be generalised by allowing more succinct descriptions of<br />systems than Kripke structures. We investigate the complexity of the<br />model checking problem when the instances of the problem consist of<br />a formula and a description of a system whose state space is at most<br />exponentially larger than the description. Based on Turing machines,<br />we define compact systems as a general formalisation of such system<br />descriptions. Examples of such compact systems are K-bounded Petri<br />nets and synchronised automata, and in these cases the well-known <br />algorithms presented in [SC85, CES86] would require exponential space in<br />term of the sizes of the system descriptions and the formulas; we present<br />polynomial space upper bounds for the model checking problem over<br />compact systems and the logics CTL and L(X,U,S). As an example of<br />an application of our general results we show that the model checking<br />problems of both the branching time temporal logic CTL and the linear<br />time temporal logics L(F) and L(X,U, S) over K-bounded Petri nets are<br />PSPACE-complete.