On-the-Fly Model Checking for Extended Action-Based Probabilistic Operators

Abstract

International audienceThe quantitative analysis of concurrent systems requires expressive and user-friendly property languages combining temporal, data-handling, and quantitative aspects. In this paper, we aim at facilitating the quantitative analysis of systems modeled as PTSs (Probabilistic Transition Systems) labeled by actions containing data values and probabilities. We propose a new regular probabilistic operator that computes the probability measure of a path specified by a generalized regular formula involving arbitrary computations on data values. This operator, which subsumes the Until operators of PCTL and their action-based counterparts, can provide useful quantitative information about paths having certain (e.g., peak) cost values. We integrated the regular probabilistic operator into MCL (Model Checking Language) and we devised an associated on-the-fly model checking method, based on a combined local resolution of linear and Boolean equation systems. We implemented the method in the EVALUATOR model checker of the CADP toolbox and experimented it on realistic PTSs modeling concurrent systems

Design and Deployment of Expressive and Correct Web of Things Applications

“…We introduce two categories of properties: generic (deadlock freedom, completeness, spurious events and progress) and application-specific properties (functional and quantitative properties). Properties are specified using the Model Checking Language (MCL) [41], a data-handling temporal language equipped with the Evaluator model checker [40] of CADP. As described earlier, by encoding in LNT an application consisting of objects and a composition expression C, we automatically derive the LTS representing its behaviour (denoted by LT S C ) using the CADP compilers.…”

“…• evaluator [40,41] is used for on-the-fly verification of a temporal property on the LTS (generated by LNT compilers in our case). It takes two inputs: an LTS encoded in one of the supported formats (BCG, LNT etc.)…”

Model checking: recent improvements and applications

“…In the paper On-the-Fly Model Checking for Extended Action-Based Probabilistic Operators by Mateescu and Requeno [ 32 ], which extends their SPIN 2016 paper [ 33 ], a new regular probabilistic operator is proposed to specify the probability measure of a path described by a generalised regular formula involving computations on data values. This operator subsumes the until operators of PCTL and their action-based counterparts.…”

Improving Graph-based methods for computing qualitative properties of markov decision processes

“…Several techniques have been proposed to improve the performance of the standard probabilistic model checking methods. These techniques are used to reduce the running time of the graph-based [20,21] or numerical computations [22][23][24] or both methods [16,25]. Although the proposed methods show promising results in the running time of probabilistic model checking, experimental results show more improvement is needed for iterative graph-based of numerical methods [26].…”