ECE Course Outline


Embedded Computing Systems (3-3-4)

ECE 3056 [min C]
Catalog Description
Algorithms and methodologies for the design of real-time, low-power embedded computing systems.
M. Wolf, Computers as Components: Principles of Embedded Computing System Design (3rd edition), Morgan Kaufmann, 2012. ISBN 0123884365, ISBN 978-0123884367 (required)

Course Objectives - As part of this course, students:

  1. learn software performance analysis and optimization including worst-case execution time. [a, c]
  2. perform real-time scheduling. [a]
  3. learn design methodologies for embedded computing systems. [c, k]
  4. design, analyze, simulate, implement, and evaluate embedded hardware and software. [c, k]
  5. use engineering techniques, skills, and tools, including software-based methods for analysis and synthesis of software and hardware [k]
  6. develop basic skills in writing laboratory reports and other documentation as part of a design methodology, including design reviews. [g]
Course Outcomes - Upon successful completion of this course, students should be able to:

  1. design and implement embedded computing systems, including software and hardware.
  2. develop verification/validation evaluation methods for these hardware and software components and analyze and interpret the resulting data.
  3. write laboratory reports and documentation conforming to technical writing standards, including widely used methodologies for design documentation.
  4. analyze the execution time of small sections of assembly language code.
  5. analyze and estimate the worst-case execution time of a program.
  6. analyze and estimate the worst-case power consumption of a program.
  7. describe CPU mechanisms that support context switching.
  8. describe the steps required to switch contexts in a CPU.
  9. compute the CPU utilization of a set of tasks.
  10. analyze the schedulability of a set of tasks using rate monotonic scheduling.
  11. design and implement software that performs a real-time task.
  12. describe an embedded computing system design using UML.
  13. decompose a set of requirements for an embedded computing system into a set of tasks.
Topical Outline
* Challenges in embedded computing: real-time, low power.
* Comparative computer organization and instruction sets.
* Hardware and software for I/O.
* Instruction-level performance analysis under the influence of pipelines, caches, and memory management. 
* Embedded computing platforms.
* Models for computer programs.
* Path-based performance analysis and optimization, worst-case execution time, average case analysis. 
* Software power analysis and optimization.
* Processes and tasks.
* Operating system structures for context switching.
* Task graph models and CPU utilization.
* Rate-monotonic and earliest deadline first scheduling.
* Embedded multiprocessors.
* Applications: automotive, multimedia, etc.