Inferring Approximated Models for Systems Engineering

Abstract

International audienceEngineering safe and reliable systems demands rigorous approaches such as formal methods, using models. Since models are not always available, one needs to infer them from software artifacts. This paper defines a new inference approach for input-output systems that is based on FSM-based testing theory. Central to the approach is the notion of initial quotient of an FSM associated with a partial characterization set that controls the precision of this approximated model. The proposed method infers a model of a system under test by building increasingly precise quotients of it using counterexamples. Various experiments demonstrate its practical usability

Integration testing of communicating systems with unknown components

“…We will develop in Section 4 how we proceed from there. In a different paper [18], we have shown how to infer a more precise model with a counterexample. This is done by extending the initial Σ set with other suffixes learnt by applying a fixpoint procedure on the observation tree extended with the counterexample.…”

Inferring Finite State Machines Without Reset Using State Identification Sequences

“…It can work with any set Z and just ensures that we characterize a state reached w.r.t. Zequivalence [12]. However, an incorrect value for n could disrupt the localizer.…”

“…To put this in perspective, we had initially tried classical inference methods [12] [16] on web applications, where a reset would typically take more than 1 minute (restarting a virtual machine and restoring data configuration), whereas interacting with it over a local network would take around 1ms per I/O pair. In such typical contexts, an inference sequence of length 10 5 would require no more time than a single reset.…”

hW-inference: A heuristic approach to retrieve models through black box testing

“…From LocW (Groz et al, 2015), it reuses the idea of characterizing sets with a W -set, as well as the general structure of the algorithm. Progressive refinement of W was taken from Petrenko et al (2014).…”

“…Some researches also try to use more recent passive inference algorithms, such as the Blue-Fringe technique, combining them with queries as in Walkinshaw et al (2007). Petrenko et al (2014) is another example of combining passive state merging with active learning. Indeed, combining monitoring (passive) with testing (active) is an active field of research especially in the field of discovering so-called software protocols, business processes or service orchestration (Bertolino et al, 2009;Dallmeier et al, 2012).…”