Completing EAST-ADL2 with MARTE for enabling scheduling analysis for automotive applications

Saoussen Anssi (1) , Sara Tucci-Pergiovanni (2) , Chokri Mraidha (3) , Arnaud Albinet (4) , François Terrier (2) , Sébastien Gérard (2)

1 Continental Automotive GmbH
2 LISE - Laboratoire d'Ingénierie dirigée par les modèles pour les Systèmes Embarqués
3 DILS (CEA, LIST) - Département Ingénierie Logiciels et Systèmes
4 Siemens VDO [Toulouse] - Siemens VDO

Automotive software systems become more and more complex presenting tougher safety requirements and tighter timing constraints. On the other hand, today, timing verification of automotive systems is considered late in the development process, mainly at the implementation stage. Consequently, this leads generally to late detection of design mistakes involving extra costs. In this paper, we propose a model driven approach to perform real-time analysis since the earliest design phases. This approach is based on the combination of two modeling languages for system design (EAST-ADL2 and MARTE) and the integration of an open source toolset for scheduling analysis, MAST. * Benefits of the proposed approach are discussed through a practical automotive case study.

Mots clés: Modeling; MARTE; Scheduling analysis; EAST- ADL2; Timing verification

Domaines: Systèmes embarqués

Saoussen Anssi, Sara Tucci-Pergiovanni, Chokri Mraidha, Arnaud Albinet, François Terrier, et al.. Completing EAST-ADL2 with MARTE for enabling scheduling analysis for automotive applications. Embedded Real Time Software & Systems (ERTS² 2010), May 2010, Toulouse, France. ⟨hal-02264387⟩