1 Embedded Systems Engineering, Department of Informatics and Mathematical Modeling, Technical University of Denmark2 Department of Informatics and Mathematical Modeling, Technical University of Denmark3 Copenhagen Center for Health Technology, Center, Technical University of Denmark
Embedded systems are now omnipresent: from cellular phones to pagers, from microwave ovens to PDAs, almost all the devices we use are controlled by embedded systems. Many embedded systems have to fulfill strict requirements in terms of performance and cost efficiency. Emerging designs are usually based on heterogeneous architectures that integrate multiple programmable processors and dedicated hardware components. New tools which extend design automation to system level have to support the integrated design of both the hardware and software components of such systems. This thesis concentrates on aspects of scheduling and communication for embedded real-time systems. Special emphasis has been placed on the impact of the communication infrastructure and protocol on the overall system performance. The scheduling and communication strategies proposed are based on an abstract graph representation which captures, at process level, both the dataflow and the flow of control. We have considered non-preemptive static cyclic scheduling and preemptive scheduling with static priorities for the scheduling of processes, while the communications are statically scheduled according to the time triggered protocol. We have developed static cyclic scheduling algorithms for time-driven systems with control and data dependencies. We show that by considering aspects of the communication protocol, significant improvements can be gained in the schedule quality. In the context of event-driven systems we have proposed a less pessimistic schedulability analysis that is able to handle both control and data dependencies. Moreover, we have provided a schedulability analysis for the time-triggered protocol, and we have proposed several optimization strategies for the synthesis of communication protocol parameters. Extensive experiments as well as real-life examples demonstrate the efficiency of our approaches.