Abstract
The emergence of wireless sensor networks (WSNs) in different and complex applications has made their focus on numerous research areas. A prominent effort focused on reconfiguration approaches that present a significant concern in sensor networks. A WSN system must be capable of modifying its behavior according to its requirements or the environment changes in ways that increase network longevity. This has prompted the development of several software solutions and mechanisms that can enable the design and development of self-adaptive WSN. However, there is still a lack of support in expressing high-level techniques which provide more abstract and reusable designs. In this sense, this paper presents a new solution for the design and development of a reconfigurable WSN system based on model driven engineering (MDE) paradigm called EARN (Energy Aware Reconfigurable Node)-MDE solution. The proposed process offers a complete methodology for the development of reconfigurable WSN from the modeling step to the final code generation. The EARN-MDE process is based on design patterns and MDE concepts. It permits the automatic generation of a high-level reconfigurable WSN model, specifically network level based reconfiguration. It offers also Model-To-Text transformations for automatically generating codes that enable the system description and analysis phase. Finally, the EARN-MDE process is tested and validated through the EARNPIPE demonstrator located at the Digital Research Center of Sfax