AN ENTERPRISE ARCHITECTURE AND MODEL DRIVEN ENGINEERING BASED APPROACH FOR SENSOR NETWORKS

Marine observatories (MO) based on sensor networks provide a continuous ocean monitoring. These sensor networks contain several kinds of sensors including acoustic hydrophones to detect and localize moving objects or animals like dolphins. In the context of marine observatories, the sensor networks provide high level services and are included in an information system to process, store and present the sensor data. This kind of system is considered as complex system and is assimilated as enterprise system with business rules and services and with several hypothesis to map these services to the distributed enterprise infrastructure. To specify, develop and deploy such systems remains a challenge to satisfy the needs, and the associated requirements, with the respect of the platform constraints. So, one of the questions is how to improve life-cycle of these systems to contribute the architecture design which is one of the sensible phase. Because this phase is the crucial one to obtain the best trade-of between the services and the infrastructure. So in this work, we try to contribute a system life-cycle based on the use of a model driven approach with an early validation phase to support ease up the development and deployment phases. The use of the models provide the facility to apply an iterative approach at system level which remains a challenge compare to the software processes. In this document, we present our approach based on an Enterprise Architecture Framework to take into account the complexity of the system. These frameworks provide the capacity to model the system on several viewpoints to express the different concerns of such systems. The choice to use an Enterprise Architecture Framework, and the associated tool ArchiMate, seems to be the most relevant due to our system features and the capacity to extend and specialize the associated tooling. The ArchiMate tooling is built on top of MDE technologies which provide facilities to extend the language definition with sensor network domain-specific concepts and constraints Thus, we propose a metamodel to define the domain concepts, and the metamodel is the support to generate a new design tool called ArchiMO. In addition, we specialize the mapping approach between the layers of the ArchiMO tool with the domain constraints to guarantee the model consistency regarding the domain. This resulting model is processed by a model compiler to generate a simulator code to achieve a simulation execution. The results of the simulation are used to analyze and validate the model of the system. After that, the iterative approach can be applied to improve the model regarding the requirements of the system, or to go forward in the development process. Our approach and tooling are demonstrated with an example from the marine observatory domain on underwater moving object localization with several acoustics sensors. This use case is used to validate our tooling to model the system, ensure consistency of the model and finally simulate the model. Through this use case, we observe that our tooling helps to reduce the complexity with the three viewpoints in the model, to improve the design activity via the domain constraints which ensure the model consistency of the Marine Observatory. As conclusion, this work aims to demonstrate that we can improve the development process of complex system based on the use of MDE technologies and a domain specific modeling language with the associated tooling. The major improvement is to provide an early validation step via models and simulation approach to consolidate the system design.