Reduction and Simplification of Explicit LTL Model Checking via an Abstraction Method

Abstract

An abstraction method developed for the explicit linear temporal logic model checking was geared towards reducing the useless part of the state space during the abstraction period. This reduces the cost during the abstraction period relative to models requiring many useless states. A dining-philosophers example comparing this abstraction method with conventional methods indicates that a large proportion of the state space has been reduced by this abstraction method. Finally, the abstract method is shown to be correct and an analysis is given to show how such a large proportion of states can be reduced.