Skip to Main content Skip to Navigation
Journal articles

Model-Driven Approach for Early Power-Aware Design Space Exploration of Embedded Systems

Abstract : Due to the growing complexity of Systems-on-Chip (SoC) and the increasing cost of their redesign and fabrication, industrials are urgently looking for design methodologies allowing them to identify issues early in the design flow and to explore the largest possible space of solutions. Several aspects should be taken into account in this context, among which power consumption is considered as a major concern. In this paper, we present a Model Driven Engineering (MDE) approach for early power-aware Design Space Exploration (DSE). This approach facilitates designers work by abstracting the energetic behavior of embedded systems through high-level models targeting an automatic generation of power-aware simulation code. It offers also the possibility to model dynamic power management aspects in order to use the corresponding generated code for DSE. This approach was implemented in the DSE toolkit TTool by integrating power concepts in its DIPLODOCUS UML profile and its simulator. This paper illustrates the proposed approach through a Software-Defined Radio (SDR) case study integrating the Dynamic Slack Reclamation (DSR) policy for dynamic power management. The processor power estimates obtained by the generated simulation code were compared to those obtained from physical implementation on the Xilinx Zynq-7000 platform. This comparison showed that our MDE approach allows to take efficient design decisions early in the design flow.
Document type :
Journal articles
Complete list of metadata
Contributor : Kathleen TORCK Connect in order to contact the contributor
Submitted on : Monday, October 25, 2021 - 4:22:42 PM
Last modification on : Wednesday, November 3, 2021 - 5:24:23 AM




Fériel Ben Abdallah, Chiraz Trabelsi, Rabie Ben Atitallah, Mourad Abed. Model-Driven Approach for Early Power-Aware Design Space Exploration of Embedded Systems. Journal of Signal Processing Systems, Springer, 2017, 87 (3), pp.271-286. ⟨10.1007/s11265-016-1144-3⟩. ⟨hal-03402454⟩



Record views