A Reasoning Method for Timed CSP Based on Constraint Solving

Abstract

Timed CSP extends CSP by introducing a capability to quantify temporal aspects of sequencing and synchronization. It is a powerful language to model real time reactive systems. However, there is no verification tool support for proving critical properties over systems modelled using Timed CSP. In this work, we construct a reasoning method using Constraint Logic Programming (CLP) as an underlying reasoning mechanism for Timed CSP. We start with encoding the semantics of Timed CSP in CLP, which allows a systematic translation of Timed CSP to CLP. Powerful constraint solver like CLP($\mathcal{R}$) is then used to prove traditional safety properties and beyond, e.g., reachability, deadlock-freeness, timewise refinement relationship, lower or upper bound of a time interval, etc. Counter-examples are generated when properties are not satisfied. Moreover, our method also handles useful extensions to Timed CSP. Finally, we demonstrate the effectiveness of our approach through case study of standard real time systems.