An extensible infrastructure for modeling and time analysis of predictable embedded systems

Efficient design of predictable systems on top of multiprocessor-based architectures is challenging. It demands an integration effort to support system models relying on Models-of- Computation (MoC) theory, supporting real-time (RT) analysis and electronic system-level (ESL) design techniques. This paper presents a SystemC-based framework for modelling and time analysis of predictable embedded systems which aims such an integration. The framework has features for system-level design and research of predictable systems. Moreover, the framework is extensible, to enable experts from different communities to explore and assess their contributions, e.g. new schedulers, schedulability analyses, and predictable platform components, without having to rely on a physical platform.