Method for automatically closing open reactive systems

Abstract

A unique process, i.e., method, is employed to automatically close an open reactive system. This is realized by employing a unique analysis technique to automatically close the open system by eliminating its interface to its environment. This allows the open system being analyzed to be comprised of either a single sequential process or a set of concurrent processes. The effect of closing the system is to make it executable and amenable to analysis. This is because the resulting closed system simulates the behavior of the original open system in its most general environment. Specifically, this is realized by employing a unique method that transforms an open system into a closed nondeterministic system such that all data values in the open system and its environment that may depend on the behavior of the environment are eliminated in the resulting closed system, and all control-flow choices in the open system and its environment that may depend on these data values are replaced by nondeterministic choices in the closed system. The reactive behavior of the open system and its environment and the resulting closed system, as well as their effect on data values that do not depend on the open system environment, are closely related. For example, every execution of the open system and its environment corresponds to an execution of the resulting closed system that exhibits the same sequence of visible operations and that preserves all data values that do not depend on the open system environment. All deadlocks and all assertion violations in a transition system of the open system and its environment that evaluate only expressions each of whose value does not depend on the open system environment are preserved in a new closed system transition system.