Proving the correctness of reactive systems using sized types

Abstract

We have designed and implemented a type-based analysis for proving some basic properties of reactive systems. The analysis manipulates rich type expressions that contain information about the sizes of recursively defined data structures. Sized types are useful for detecting deadlocks, nontermination, and other errors in embedded programs. To establish the soundness of the analysis we have developed an appropriate semantic model of sized types.