Digital Signal Processing

Digital Signal Processing (DSP) deals with the manipulation and shaping of signals whose values are only defined at discrete instants in time. Discrete signals are a natural consequence of measuring physical and process variables at discrete intervals in time. Continuous signals are often sampled by Analog-to-Digital Converters (A/D) for direct storage in a computer. Signals stored in a computer can be processed for many practical applications. Examples include speech synthesis, speech coding, digital audio and video, digital control of industrial processes, digital filtering for shaping signal characteristics including noise suppression, etc.

The DSP certificate program will introduce the participants to basic theoretical and practical concepts of DSP. Applications of digital filtering, speech processing, real-time processing, and FFT analysis will be discussed. Case studies of typical DSP applications will be presented. (12 credit hours) 

Coursework

Please choose four courses to complete the required 12 credit hours.
ECE 512Analog Filter Design3
ECE 565Digital Control Systems3
ECE 576Information Engineering3
ECE 579Intelligent Systems3
ECE 580Digital Signal Processing3
ECE 5802Multirate Sig Proc w/Appl3
ECE 581Arch for Digital Signal Proc3
ECE 582Intro to Statistical DSP3
ECE 583Artificial Neural Networks3
ECE 584Speech Processes3
ECE 589Multidimen Digital Signal Proc3

ECE 512     Analog Filter Design     3 Credit Hours

This course addresses the analysis and design of continuous time (analog) and switched-capacitor filters. Students will analyze and design filters. Effect of tolerances of circuit elements on the performance of the circuit behavior will be analyzed. Students will use simulation tools to design filters and verify circuit performance. Three lecture hours per week.

Restriction(s):
Can enroll if Class is Graduate
Can enroll if Major is Electrical Engineering, Computer Engineering

ECE 565     Digital Control Systems     3 Credit Hours

Mathematical representation of digital control systems; z-transform and difference equations; classical and state space methods of analysis and design; direct digital control of industrial processes.

Restriction(s):
Can enroll if Class is Graduate

ECE 576     Information Engineering     3 Credit Hours

This course will cover fundamental concepts of information engineering, including theoretical concepts of how information is measured and transmitted, how information is structured and stored, how information can be compressed and decompressed, and information networks such as social networks, affiliation networks and online networks, mathematical theories of information networks. Information engineering applications will be discussed. Three lecture hours per week.

Restriction(s):
Can enroll if Class is Graduate
Can enroll if Level is Doctorate or Rackham or Graduate
Can enroll if Major is Computer Engineering, Software Engineering, Electrical Engineering, Computer & Information Science

ECE 579     Intelligent Systems     3 Credit Hours

Representative topics include: Intelligent systems design, training and evaluation, decision trees, Bayesian learning, reinforcement learning. A project will be required.

Restriction(s):
Can enroll if Level is Rackham or Graduate or Doctorate
Can enroll if Major is Computer & Information Science, Software Engineering, Electrical Engineering, Computer Engineering

ECE 580     Digital Signal Processing     3 Credit Hours

This course addresses the analysis and design of discrete ?time signals and systems. Students will become familiar with the mathematical tools needed for digital signal processing such as the Fourier transform, frequency response, the sampling theorem, and z-transform method. Topics covered will include the design of digital filters (IIR and FIR filters), characteristics of analog-to-digital and digital-to-analog converters, the spectral analysis of signals, and discrete filters. Three lecture hours per week.

Restriction(s):
Can enroll if Class is Graduate

ECE 5802     Multirate Sig Proc w/Appl     3 Credit Hours

This course provides an introduction to multirate digital signal processing with application in different fields of engineering, with a focus on the presentation of the theoretical foundation for all aspects of multirate digital signal processing. The course examines modern applications of multirate digital signal processing including the design of multirate filter banks, using the wavelets transforms to efficiently encode signals for compression purposes, spectral analysis and synthesis of signals. Students will apply software tools to analyze, design and simulate multirate digital signal processing systems. Three lecture hours per week.

Prerequisite(s): ECE 580

Restriction(s):
Can enroll if Level is Graduate or Rackham or Doctorate
Can enroll if Major is Mechanical Engineering, Automotive Systems Engineering, Computer Engineering, Information Sys Engineering, Software Engineering, Industrial & Systems Engin, Electrical Engineering, Engineering Management

ECE 581     Arch for Digital Signal Proc     3 Credit Hours

This course introduces the architectural fundamentals and features of programmable digital signal processors. Numeric representations and arithmetic concepts are discussed, which include fixed-point and floating-point representation of numbers, native data word width, and IEE-754 floating-point representation. Memory architecture and memory structures of digital signal processors are examined. Programming concepts for DSP processors such as addressing, instruction set, execution control, pipelining, parallel processing and peripherals are discussed. Finally, students will work on related applications employing digital signal processors such as speech processing, instrumentation, and image processing. Three lecture hours per week.

Prerequisite(s): ECE 580

Restriction(s):
Can enroll if Class is Graduate
Can enroll if Major is Computer & Information Science, Electrical Engineering, Computer Engineering

ECE 582     Intro to Statistical DSP     3 Credit Hours

Review of discrete-time signals and systems, introduction of discrete-time random signals and variables, linear signal models, nonparametric power spectrum estimation, least-squares filtering and prediction, signal modeling and parametric spectral estimation, selected topics. (W).

Prerequisite(s): ECE 580*

Restriction(s):
Can enroll if Class is Graduate
Can enroll if Major is Electrical Engineering

ECE 583     Artificial Neural Networks     3 Credit Hours

Students will gain an understanding of the language, formalism, and methods of artificial neural networks. The student will learn how to mathematically pose the machine learning problems of function approximation/supervised learning, associative memory and self-organization, and analytically derive some well-known learning rules, including backprop. The course will cover computer simulations of various neural network models and simulations. Three lecture hours per week.

Restriction(s):
Can enroll if Class is Graduate
Can enroll if Level is Graduate or Rackham or Doctorate
Can enroll if Major is Computer Engineering, Electrical Engineering, Computer & Information Science, Software Engineering

ECE 584     Speech Processes     3 Credit Hours

The course introduces the fundamentals of speech processing using digital signal processing methods and techniques. How speech is produced from the human vocal system and the different types of basic speech sound components is addressed, followed by methods to analyze speech signals in both the time domain and frequency domain. Applications of speech processing are also presented. Possible applications covered include speech synthesis, speech coding and speech recognition. A team-based term project may be required. Three lecture hours per week.

Restriction(s):
Can enroll if Class is Graduate

ECE 589     Multidimen Digital Signal Proc     3 Credit Hours

Topics include multidimensional signal analysis methodologies, signal representation, 2-D FIR filter, 2-D recursive systems and IIR filters, spectral estimation and methods, multidimensional signal restoration applications in 2-D and 3-D image processing, reconstruction, and feature estimation. Three lecture hours per week.

Prerequisite(s): ECE 580

 
*

An asterisk denotes that a course may be taken concurrently.

Frequency of Offering

The following abbreviations are used to denote the frequency of offering: (F) fall term; (W) winter term; (S) summer term; (F, W) fall and winter terms; (YR) once a year; (AY) alternating years; (OC) offered occasionally