A review on specifying software architectures using extended automata-based models

Abstract

Applying an appropriate formal model to specify software architecture makes a reliable foundation to formally verify non-functional properties and therefore, leads to early detection of defects. In this paper we make a comparison between automata-based models and evaluate their abilities to model different aspects of components interaction in software architectures. We try to use Team automata as a middleware to formally specify well-known architectural descriptions in UML2.0. A Limitation of current automata models, so called "actions interleaving" is also discussed and some approaches to overcome this limitation described. © Springer-Verlag Berlin Heidelberg 2007.