A Method for Improving the Verification and Validation of Systems by the Combined Use of Simulation and Formal Methods

Abstract

Verification and Validation (V&V) of Systems is an important process in the development of systems, in order to ensure that they are reliable and operational. Among methods of V&V, there are two that seem to be opposite to each other: simulation, which is empirical, and formal verification, which is comprehensive. Moreover, simulation and formal verification propose many different formalisms, increasing the gap between them. But, jointly used, these two powerful tools allow making a more efficient verification, increasing the confidence we can put in the verified systems. The main problem is how we can combine their use and how we can reduce the gap created by the nature of both of them. This paper presents guidelines and a general approach in order to use simulation, and especially discrete-event simulation, on a model specified in a verifiable formal language.

Formal verification with HiLLS-specified models: A further step in multi-analysis modeling of complex systems

“…From a more general perspective, this work differs from works related to combining multiple analysis methods in that those related works usually realize pair wise integrations of their corresponding supported methodologies (under usual circumstances, transformation rules are defined such that one model expressed in the source formalism is given as input and a model specified in the target formalism is obtained as output). Examples of such combined use of simulation and formal analysis include [24][25][26][27]. Similarly, several proposals have been made to bridge the gap between enactment and formal methods [28][29][30][31], while some efforts to achieve pair wise integration of simulation and enactment are reported in [32][33][34][35].…”

Integrated Framework for Model-Driven Systems Engineering: A Research Roadmap

“…In recent decades, there have been considerable research efforts to establish transformations between models in separate planes. For instance, research efforts to combine DEVS simulation models and FM include [20,34,36,40]. Also, combination of FM and software models have been proposed in [6,21,22] while [26,27,38] present some of the efforts to integrate simulation and software development.…”