Distributed control for reconfigurable FPGA systems: a high-level design approach

Due to their exponential complexity, designing adaptation control for Reconfigurable Systems-on-Chip (RSoC) is becoming one of the most challenging tasks, resulting in longer design cycles and increased time-to-market. This paper addresses this issue and proposes a novel adaptation control design approach for FPGA-based reconfigurable systems aiming to increase design productivity. This approach combines control distribution and high-level modeling in order to decrease design complexity and enhance design reuse and scalability. Control distribution is based on allocating local control aspects (monitoring, decision and reconfiguration) to distributed controllers, while respecting global system constraints/objectives using a coordinator. High-level modeling makes use of Model-Driven Engineering and the MARTE (Modeling and Analysis of Real-Time and Embedded Systems) standard in order to move from high level models to automatic code generation, which significantly simplifies the control design. The proposed design approach is integrated in a model-driven RSoC design flow and allows to model adaptation aspects at different design levels: application, architecture, allocation and deployment, which allows to target a wide range of control requirements. In order to validate our approach, a video processing application was implemented on a reconfigurable system that contained four distributed hardware controllers.