A model driven based approach for the design of dynamically reconfigurable systems : from MARTE to RECOMARTE

The works presented in this dissertation propose a co-design methodology of dynamically reconfigurable systems based on FPGA. Our methodology is based on the Engineering Model Driven approach (MDE) and the models specification is done in UML MARTE (Modeling and Analysis of Real-Time and Embedded Systems) profile. It aims to ensure flexibility, reusability and automation to facilitate the work of designer and improve his productivity. The first contribution related to this thesis is identifying parts of dynamically reconfigurable FPGA that can be modeled at high abstraction levels. So, we defined a design flow based on the MDE to ensure the automation of code generation. Using this flow, several models are created mainly through MARTE profile concepts. However, the modeling concepts of dynamic reconfiguration on FPGAs required extensions in MARTE. Thus, we identified the missing concepts to be integrated in a new profile that extends MARTE called RecoMARTE. The second contribution allows the chain automation and experimental validation. To integrate our design flow and to automate code generation, a processing chain was used. The final model resulting from the proposed MARTE-based design flow is given as input to this chain.We thereby move from MARTE/RecoMARTE models to an intermediate description according to the IP-XACT standard to finally generate files describing the complete system in the Xilinx XPS environment. This automation allows to accelerate the design phase and avoid errors due to the direct manipulation of these details. Finally, the proposed MARTE-based design flow and transformation chain were used for an image processing system design, which showed the benefits of our contributions in terms of design reusability and automation.

Data and Resources

Additional Info

Field Value
Source https://theses.hal.science/tel-00998248
Author Cherif, Sana
Maintainer CCSD
Last Updated May 5, 2026, 09:56 (UTC)
Created May 5, 2026, 09:56 (UTC)
Identifier tel-00998248
Language fr
Rights https://about.hal.science/hal-authorisation-v1/
contributor Laboratoire d'Informatique Fondamentale de Lille (LIFL) ; Université de Lille, Sciences et Technologies-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lille, Sciences Humaines et Sociales-Centre National de la Recherche Scientifique (CNRS)
creator Cherif, Sana
date 2013-12-19T00:00:00
harvest_object_id f5168657-f038-4ccc-ba7a-29539b69c0ef
harvest_source_id 3374d638-d20b-4672-ba96-a23232d55657
harvest_source_title test moissonnage SELUNE
metadata_modified 2025-03-23T00:00:00
set_spec type:THESE