ARTIST 嵌入式系统设计暑期讲习班


首页 课程 日程 注册 地点 宾馆 主办


  • Professor Luis Almeida (University of Porto, Portugal)
  • Real-Time Communication in Embedded Systems
  • 摘要:

    Distributed real-time systems are becoming pervasive, either in process control, factory automation and more recently, embedded systems.

    This course will present an introduction to computer networks from a real-time systems perspective. The lectures include an initial presentation of general concepts in networks and then focus on the physical and data link layers, devoting particular attention to the medium access control.

    Then, the course will address the issue of traffic scheduling and its relationship with medium access control, showing typical schedulability analysis that can be used to derive a priori guarantees of traffic timeliness. A few paradigmatic protocols will then be presented and analyzed, including a reference to the growing interest on wireless communication.

    Finally, the course will address a few related on-going research efforts, mainly towards flexible real-time communication for adaptive and reconfigurable systems.

  • Professor Karl-Erik Arzen (Lund University, Sweden)
  • Control for Embedded Systems
  • 摘要:

    The aim of the course is to give an overview of embedded control systems and of the use of control techniques in computer software systems. The course is intended for an audience having their main background in computer science and engineering. The course consists of five modules, each consisting of two 60 minute hours, with a 15 minute break in between.

    Course Modules:

    - Introduction to Feedback Control The role of feedback. Models and linearization. Stability. State-space and input-output models. Pole-placement. State-feedback and observers. Feedforward.

    - Computer Implementation of Control Systems Discretization of continuous-time control designs. Discrete-time control design. Aliasing. Anti-windup. Mode-handling. Numerics. PID control example. Task models for control.

    - Interaction between Control and Scheduling Interaction between control design and computer implementation. Temporal robustness. The effects of latencies and jitter on control performance. The Jitter Margin. The Control Server Model. Networked Embedded Control.

    - Co-Design Tools TrueTime – co-simulation of real-time kernels, networks, and continuous plants. Jitterbug – analytical temporal robustness evaluation of control loops. Several examples and demos will be shown.

    - Control of Computer Systems Examples of feedback in computer and communication systems. Queue-length control. Control of web-servers. Feedback scheduling in control systems. Feedback-based resource management.

  • Professor Kim Guldstrand Larsen (University of Aalborg, Denmark)
  • Validation, Synthesis and Performance Evaluation of Embedded Systems
  • 摘要:

    Model-driven development is a key to dealing with the increasing complexity of embedded systems, while reducing the time and cost to market. The use of models should permit early assessment of the functional correctness of a given design as well as requirements for resources (e.g. energy, memory, and bandwidth) and real-time and performance guarantees. Thus, there is a need for quantitative models allowing for timed, stochastic and hybrid phenomena to be modeled and analyzed.

    UPPAAL is a tool for modeling, simulating and verifying real-time and hybrid systems, developed collaboratively by Deparment of Computer Science at Aalborg University and Department of Computer Systems at Uppsala University since the beginning of 1995 (see UPPAAL and the branches CORA and TIGA provide an integrated tool environment for modeling, validation, verification and synthesis of real-time systems modeled as networks timed automata, extended with data types and user-defined functions. The lectures will provide details on the expressive power of timed automata in relationship to embedded systems as well as details on the power and working of the UPPAAL verification engine.

    The lectures will also introduce a new highly efficient statistical model checking engine of UPPAAL settling soft-real time performance properties of the type "within T time-units a given request R will have been granted with probability at least P" with a desired level of confidence.

    During the lectures the demonstration and application of the UPPAAL tool suite will be given on a number of practical and industrial cases, including schedulability analysis and WCET analysis.

  • Professor Jan Madsen (Technical University of Denmark)
  • Platform-Based Design: From Multi-Core Platforms to Biochips and beyond
  • 摘要:

    One of the challenges in modern embedded system design is to map the application onto a platform such that essential requirements are met. In order to do so at an early stage in the design process, where not all parts have been implemented or even designed, a system-level model of the application executing on the platform is needed. This model should allow for an accurate modeling of the global performance of the system, including the interrelationships among the diverse processing elements, physical interfaces and inter-connections. This course gives an introduction to the problem of mapping applications onto platforms.

    The course is split in two parts:

    Part 1: Multi-core platforms. The process of mapping covers the allocation of tasks to processors of the platform and the definition of their execution order, i.e. the task scheduling. The course will focus on task scheduling for parallel systems. It will cover basic architectures for multi-core platforms and how to model these, as well as how to model the application as a parallel program. The course will cover both basic scheduling algorithms (handling static scheduling) and more advanced algorithms, which are able to handle consequences of the, often complex, communication structures of the platform.

    Part 2: Biochips. Microfluidic biochips are replacing the conventional biochemical analyzers, and are able to integrate on-chip all the necessary functions for biochemical analysis. There are several types of microfluidic biochips, each having advantages and limitations. In flow-based biochips the microfluidic channel circuitry on the chip is equipped with chip-integrated microvalves that are used to manipulate the on-chip fluid flow. By combining several microvalves, more complex units like mixers, micropumps, multiplexers etc. can be built up, with hundreds of units being accommodated on one single chip. In droplet-based biochips, the liquid is manipulated as discrete droplets on an electrode array. For both types of biochip, the synthesis process, starting from a biochemical application and a given biochip architecture, determines the resource allocation, binding, scheduling and placement of the application operations, resembling the mapping process for multi-core platforms. The course will illustrate how techniques and methods from multi-core platforms can be used to solve synthesis and optimization problems of biochips.

  • Prof. Dr. Peter Marwedel (Technical University of Dortmund)
  • Embedded System Foundations of Cyber-Physical Systems
  • 摘要:

    Cyber-Physical Systems (CPS) are integrations of computation and physical processes [Edward Lee]. Information processing systems in such integrations are called embedded systems. The tutorial will start by describing the key characteristics of embedded systems as well as requirements for their specification techniques and modeling. Different models of computation will be presented. The second part of the tutorial will explain the impact of the integration into cyber-physical systems on hardware and software components. The third part will focus on the mapping of applications onto embedded platforms and techniques for the evaluation of the resulting designs. The fourth and final part will focus on a treatment of optimizations aiming at a reduction of the energy consumption and the worst-case execution time (WCET).

    Overall, the tutorial will be structured as follows: Introduction, motivation, characteristics and modeling techniques Impact of the integration of information processing with physics on hardware and software components Mapping of applications onto platforms and how to evaluate the resulting designs Optimizations minimizing energy consumption and worst case execution time (WCET)

    The presentation will provide a context for the other presentations.