From Model-based Design to Virtual Prototypes for Automotive Applications

In this paper, we present a new design methodology for automotive applications, combining the strength of model-based design using MATLAB/Simulink and virtual prototyping using SystemC. The design flow starts from an automotive application modeled in MATLAB/Simulink. By including a vehicle model, testing and debugging of the desired application is possible. In a first step, the application model is automatically transformed into SystemC code. In a second step, the ECU (Electronic Control Unit) architecture including control units and communication buses of the vehicle is also modeled in SystemC. While the application model defines the functional property of the envisioned algorithm, the ECU architecture model is responsible for modeling non-functional properties, e.g., time, area and power consumption. Finally, the SystemC application model is related to the ECU architecture model by mapping activations of the application to the modeled control units and buses. In that way, non-functional properties can be evaluated by the help of the SystemC simulation kernel. Furthermore, the flexibility of the proposed approach allows for assessing the effect of different design decisions in early design phases, as the entire ECU architecture is modeled in SystemC. Hence, partitioning the application model onto many ECUs or multiple processors inside ECUs can be done easily by only changing a single configuration file. We will demonstrate the benefits of the proposed approach using a brake-by-wire application mapped onto an ECU architecture based on a FlexRay bus system.

[1]  Ying Cai,et al.  A Transformational approach to constraint relaxation of a time-driven simulation model , 2000, ISSS '00.

[2]  Ed F. Deprettere,et al.  An Approach for Quantitative Analysis of Application-Specific Dataflow Architectures , 1997, ASAP.

[3]  Luigi Carro,et al.  Simulink-Based MPSoC Design Flow: Case Study of Motion-JPEG and H.264 , 2007, 2007 44th ACM/IEEE Design Automation Conference.

[4]  Christian Haubelt,et al.  Efficient Representation and Simulation of Model-Based Designs , 2006, FDL.

[5]  Alexandru Turjan,et al.  System design using Khan process networks: the Compaan/Laura approach , 2004, Proceedings Design, Automation and Test in Europe Conference and Exhibition.

[6]  Dieter Monjau,et al.  Workshop "Methoden und Beschreibungssprachen zur Modellierung und Verifikation von Schaltungen und Systemen" , 2001, Informationstechnik Tech. Inform..

[7]  Stavros Tripakis,et al.  From simulink to SCADE/lustre to TTA: a layered approach for distributed embedded applications , 2003, LCTES '03.

[8]  Christian Haubelt,et al.  Semi-Automatic Generation of mixed Hardware/Software Prototypes from Simulink Models , 2008, MBMV.

[9]  Nacer-Eddine Zergainoh,et al.  Simulink-based MPSoC Design: New Approach to Bridge the Gap between Algorithm and Architecture Design , 2007, IEEE Computer Society Annual Symposium on VLSI (ISVLSI '07).

[10]  Christian Haubelt,et al.  Task-accurate performance modeling in SystemC for real-time multi-processor architectures , 2006, Proceedings of the Design Automation & Test in Europe Conference.