Statistical model checking of e-motions domain-specific modeling languages

Abstract

Domain experts may use novel tools that allow them to design and model their systems in a notation very close to the domain problem. However, the use of tools for the statistical analysis of stochastic systems requires software engineers to carefully specify such systems in low level and specific languages. In this work we line up both scenarios, specific domain modeling and statistical analysis. Specifically, we have extended the e-Motions system, a framework to develop real-time domain-specific languages where the behavior is specified in a natural way by in-place transformation rules, to support the statistical analysis of systems defined using it. We discuss how restricted e-Motions systems are used to produce Maude corresponding specifications, using a model transformation from e-Motions to Maude, which comply with the restrictions of the VeStA tool, and which can therefore be used to perform statistical analysis on the stochastic systems thus generated. We illustrate our approach with a very simple messaging distributed system.