Linköping University: Students Alumni Trade and Industry/Society Internal Search

Center for Embedded Systems Engineering

The Center for Embedded Systems Engineering (CESE), conducts research in the area of real-time systems engineering. The center consists currently of two research laboratories at the Department of Computer and Information Science: ESLAB and RTSLAB (Real-Time Systems Laboratory)

Introduction

The Department of Computer and Information Science (IDA) at Linköping University is nationally a significant contributor to research in real-time and embedded systems. To further increase the competence in this field, two research groups that are active in the real-time systems area have decided to form a joint "Center for Embedded Systems Engineering" at the department. This effort complements others at Linköping University (such as ECSEL) through its strong focus on real- time systems, embedded systems, and hardware/software integration.

The research in the center will be industrially relevant, and conducted in cooperation with industry and other academic nodes in the ARTES network.

Areas of Interest

The prime purpose of CESE is to develop engineering techniques for time and safety critical embedded systems. Typical characteristics of such systems are that they must always behave correctly, both in terms of function and timing, which means that extensive validation and verification is needed. A central idea is to try to address the issues earlier in the design process than is currently done, thereby identifying and avoiding possible pitfalls before these can cause serious problems. This requires the development of analysis models which predict the outcome of later design steps, and refinement techniques that transform a specification into an implementation in a way that guarantees its correctness. More specifically, the center will concentrate its initial efforts in the following areas:
  • Systematic design with a formal basis. Determining through ordinary testing that an implementation meets its specification is a very tedious task. An alternative approach is instead to refine the specification into an implementation, by making smaller transformations that can be shown formally to preserve the system's behaviour. A systematic strategy based on these ideas will be developed, and also form the basis for synthesis tools, which do parts of the refinement and verification automatically.
  • Performance and resource analysis. A system's timing behaviour, and its ability to meet timing constraints, depend to a large degree on how it is implemented, i.e. what resources it uses. To make assessments early in the design process of the systems temporal behaviour, its resource needs must be analyzed. This includes determining the execution time of tasks on different processors, but also analyzing how hardware resources (e.g. processors, buses) are time-shared by different tasks in the system.
  • Hardware/software co-design. Since it is the combined effect of software and hardware that determines the final implementation's characteristics, it is natural to closely coordinate the development of these parts. Also, the borderline between hardware and software implementations is becoming more and more diffuse, since it is nowadays possible to automatically synthesize application-specific circuits from program-like descriptions. This project will address system-level design and synthesis of mixed hardware/software systems, with special emphasis on real-time issues.