Feedback Control Learning Model for QoS, Power & Performance Management of Reconfigurable Embedded Systems
Résumé
This paper focuses on a usually ignored issue, the autocon guration modelling and decision in the context of reconfigurable embedded systems. We propose an original and generic method based on control theory including stability analysis. Our approach addresses the question of local vs global reconguration decision at hardware, software and RTOS levels. We tackle the run-time uncertainty conditions with a learning system model that balances tradeoffs between accuracy and complexity using sensors and run-time light estimators based on signal processing theory. Then we figure out how our method can be applied to a smart camera.
Domaines
Architectures Matérielles [cs.AR]
Loading...