HLS-based Accelerated Simulation of Large Scale Cyber-Physical Systems on FPGAs - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année :

HLS-based Accelerated Simulation of Large Scale Cyber-Physical Systems on FPGAs

(1, 2) , (3, 4) , (3, 5) , (2) , (2)
1
2
3
4
5

Résumé

The design of cyber-physical systems remains challenging because of their highly heterogeneous nature that makes modeling, design and analysis hard. Despite extensive work in model-based approaches, few unified simulation tools are available today for such systems. This paper proposes a simulation strategy that benefits from the characteristics of recent FPGA platforms and advances in the high-level synthesis tools. Our proposal consists in using these tools to build a cyber-physical system simulator running at high-speed on a FPGA; in this view, high-level synthesis is used not only in the traditional prototyping phase of the embedded systems, but also to synthesize its physical environment, which is jointly simulated on the FPGA. Our paper proposes a case study illustrating this approach: the simulation of the automatic identification system required in maritime communications. The simulation executed on the latest FPGA generation is accelerated by a factor x654 compared to software alternatives demonstrating that FPGAs exhibit appealing characteristics for such simulations.
Fichier principal
Vignette du fichier
_Newcas_2022__HLS_based_Accelerated_Simulation_of_Large_Scale_Cyber_Physical_Systems_on_FPGAs.pdf (834.36 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03839510 , version 1 (04-11-2022)

Identifiants

Citer

Maelic Louart, Jean-Christophe Le Lann, Frédéric Le Roy, Abdel Boudraa, Jean-Jacques Szkolnik. HLS-based Accelerated Simulation of Large Scale Cyber-Physical Systems on FPGAs. IEEE International NEWCAS Conference, Jul 2022, Montreal, Canada. ⟨10.1109/NEWCAS52662.2022.9842250⟩. ⟨hal-03839510⟩
0 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook Twitter LinkedIn More