Abstract
(Englisch)
|
Embedded systems represent by far the largest application area for IT systems. An embedded systems is typically composed of three components: data acquisition, data processing and action generation. All three components strongly influence the overall behaviour of the system. The target system in this project is a mobile system with embedded controllers, distributed and heterogeneous sensors, acting in a dynamically changing physical environment. The objectives of the project are to develop new paradigms and methods for modelling, verification and design of embedded control systems with distributed sensing, which include explicit handling of resource constraints and uncertainty. The focus will be on mobile systems where particular attention is needed for management of structural and real-time complexity.
Objectives:
The main driver of this proposal is the development of new paradigms and methods for control design of embedded systems with distributed sensing, limited communication and computational resources. This includes rigorous methods for modeling, analysis, design and verification of control systems with heterogeneous sensors. Although such systems are omnipresent, the proposal is focused on mobile systems, which provide a particularly challenging benchmark since they operate in an uncertain and changing environment. The new methodologies will deal with control design under various constraints, such as signal quantisation, bandwidth limitation, distributed sensing and computation, and real-time scheduling. The project will add to the overall scientific thrust on handling information complexity in systemdesign.
Work description:
The purpose of the project is to develop methodologies and unifying principles for the rational design of embedded systems that have distributed and heterogenous sensors and operate in an uncertain and changing environment. This implies that fundamental issues in the design of such embedded systems include modelling, verification, data processing from distributed and heterogenous sensors, as well as design of concurrent controllers under sensory and communication constraints. These research problems present the core of this project and the workplan is accordingly organized into four workpackages:WP1. Tools for abstraction and analysis of complex systems; WP2. Design of control systems under communication and computation constraints; WP3. Real-time processing of distributed sensory data; WP4. Implementation and validation on a mobile testbed. Although the case study in WP4 (mainly partners 4 and 5) will provide a concrete example, the main driver in this project is to further the state of art in the research of the above mentioned issues. The main purpose of WP1 is to provide a methodology for modelling and verification of complex, hybrid and quantised systems. Partner 2, in collaboration with Partners 1 and 3, will contribute basic knowledge in modelling and analyzing these systems as hybridcontrol systems, with a robust and scalable architecture. In WP2, Partners 3, 2and 1 will contribute basic knowledge and techniques in design of distributed and concurrent controllers under sensory and communication constraints. In WP3, Partners 1 and 2 will provide methods for handling computationally demanding sensing tasks and non linear filtering algorithms.In WP4 Partner 4 and 5 will develop software tools for the experimental evaluation of the techniques on a mobile system testbed.
Milestones:
The project will in particular deliver basic methods for modelling and design of embedded control systems, new methods for data processing from distributed and heterogenous sensors as well as design of concurrent controllers under sensory and communication constraints. The methods proposed will be tested in a concrete platform.
|
