Embedded systems are dedicated computing and control systems embedded into a technical environment. They execute functions in response to specific input stimuli usually within predefined time windows.

A main characteristic of embedded systems is their heterogeneous composition. Embedded systems may be composed of dedicated or weakly programmable hardware components, predesigned programmable components (software), communication networks and analogue interfaces. An embedded system is often tightly integrated with sensors and actuators communicating with the environment.

Currently, most embedded systems are designed manually. However, the growing complexity of these systems makes CAD support for the development process, life-cycle and maintenance indispensable. CAD can potentially reduce the design intervals provided proper methodology is used to enable fast prototyping and to design components in parallel. Finally, in many safety critical applications, the correctness of the realization with respect to specified real-time properties must be proved, for example using verification methods.

The success of computer aided design methods for digital integrated circuits is partly founded on the development of a design methodology based upon abstraction. Here methodologies linking different levels of abstraction have been developed for which synthesis and optimization algorithms provide a well tested path to implementation. Due to their complex heterogenous nature, pragmatic methodologies for embedded systems have not yet been developed. This heterogeneity is reflected in the huge amount of conceptually different models and methods. Despite of the high expectations, many fundamental problems are not solved yet.

This seminar was devoted to models and methods for the design of embedded systems. In particular, the presentations and discussions covered the following topics:

• Models: The relationships between the different models of computation, the hardware software dependencies in these models. The search for a sound common model for which useful design methods can be applied.
• Optimization methods: Methods for updating an implementation as a result of a change in the specification (redesign) and for hardware-softwarecodesign (partitioning, optimization, real-time scheduling).
• Interfaces and communication: Efficient concepts for linking hardware and software subsystems, formal specification and verification of interfaces between tightly coupled components and the communication between distributed subsystems.
• Performance simulation and estimation: Models and methods to estimate dynamic properties of embedded systems, i.e. communication traffic and execution profiles.
• Verification, validation and emulation: Target architecture, functionality, timing, prototyping for hardware and software including FPGA-emulators, co-simulation and co-verification of hardware and software.
• Synthesis and partitioning: Regularity, data dependency and coprocessor extraction, knowledge based partitioning, cost functions for hardware, software and communication.
• Applications: Software acceleration, coprocessors, instruction set optimization and power consumption.

The organizers would like to thank everyone who has helped to make the seminar a great success.

Creative Commons BY 3.0 DE