Supporting Simulation of Systems-of-Systems Software Architectures by a Model-Driven Derivation of a Stimulus Generator

Valdemar Vicente Graciano NetoCarlos Eduardo Barros PaesFlávio OquendoElisa Yumi Nakagawa

Systems-of-Systems (SoS) often support critical domains such as airtraffic control, and emergency and crisis response. Hence, their software architectures must be validated, guaranteeing that they conform to their specification. However, SoS exhibit dynamic properties that bring difficulties to a static validation. In this direction, simulations can aid by offering a dynamic view for software architecture specifications of SoS. However, to be reliable, simulation models must reproduce the real conditions of the surrounding environment in which it will be deployed. Conversely, manually stimulating simulations is tiring, repetitive, and error-prone. Then, we explored the possibility of automatically deriving a 'stimulus generator', that continuously emits stimulus for the entities being simulated, supporting an early identification of failures, and enabling correction before the deployment. In this direction, we present our method to automatically derive a stimulus generator from a software architecture description of SoS. The main contribution of this paper is providing such method, suppressing the necessity of manually coding it. We evaluated our proposal with an example of a real Flood Monitoring SoS. Preliminary results point out that the stimulus generator automatically produced is reliable, emitting the expected outputs, and suitably triggering the simulation.

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: