Extending Refusal Testing by Stochastic Refusals for Testing Non-deterministic Systems

TL;DR

This paper introduces a formal testing framework using stochastic refusal graphs to validate non-deterministic stochastic systems, automating tester generation and demonstrating practical application.

Contribution

It presents a novel approach for testing stochastic systems with SRGs, including automatic tester generation from MLSTS models.

Findings

Framework for generating SRGs from MLSTS models

Automatic canonical tester generation from stochastic refusal graphs

Practical implementation demonstrated with an example

Abstract

Testing is a validation activity used to check the system's correctness with respect to the specification. In this context,test based on refusals is studied in theory and tools are effectively constructed. This paper addresses,a formal testing based on stochastic refusals graphs (SRG) in order to test stochastic system represented by maximality-based labeled stochastic transition systems (MLSTS). First, we propose a framework to generate SRGs from MLSTSs. Second,we present a new technique to generate automatically a canonical tester from stochastic refusal graph and conformance relation confSRG. Finally, implementation is proposed and the application of our approach is shown by an example.