Exploring the Interaction between Functional Performance and Scheduling Abstractions
Posted: Wed Jun 07, 2017
Title: Exploring the Interaction between Functional Performance and Scheduling Abstractions
Authors:
Paolo Pazzaglia, Alessandro Biondi, Marco Di Natale, Giorgio Buttazzo and Matteo Secchiari (Scuola Superiore Sant’Anna, Pisa, Italy)
Abstract: Several schedulability analysis models and results are based on the assumption that the functional performance of the system can be represented (abstracted) by a very simple task model with release and completion time constraints. The hard deadline model is probably the best known example. However, several real applications, such as engine control challenge both the task activation and execution time model, and also the deadline abstraction and the hard schedulability assumptions. Our project on the development of a Simulink based co-simulation environment aims at a more detailed exploration of the impact of (possibly late) response time on the performance of complex control functions. We describe not only our cosimulation framework, but also additional process steps and challenges that relate to its use.
Attached paper:
Authors:
Paolo Pazzaglia, Alessandro Biondi, Marco Di Natale, Giorgio Buttazzo and Matteo Secchiari (Scuola Superiore Sant’Anna, Pisa, Italy)
Abstract: Several schedulability analysis models and results are based on the assumption that the functional performance of the system can be represented (abstracted) by a very simple task model with release and completion time constraints. The hard deadline model is probably the best known example. However, several real applications, such as engine control challenge both the task activation and execution time model, and also the deadline abstraction and the hard schedulability assumptions. Our project on the development of a Simulink based co-simulation environment aims at a more detailed exploration of the impact of (possibly late) response time on the performance of complex control functions. We describe not only our cosimulation framework, but also additional process steps and challenges that relate to its use.
Attached paper: