ELEC 532 - Spectral Analysis - Syllabus

Vocals

Prof. Richard G. Baraniuk

Time and Place

Tuesday/Thursday from 1:00 to 2:20pm CST in Physics Lab 210

Backing Vocals and Stage Manager

Ramesh "Neelsh" Neelamani
DH 2121, 713.348.3230, neelsh@rice.edu

Office Hours

RN: by appointment

Roadie

Justin Romberg

Textbook(s) - Both of these books are optional and on reserve in the library

S. M. Kay, Modern Spectral Estimation: Theory and Application, Prentice Hall, 1988
(almost out of print).

S. L. Marple, Digital Spectral Analysis with Applications, Prentice Hall, 1987.

Recommended Further Reading

The Student Edition of MATLAB, Prentice Hall

The Matlab Primer

Prerequisites

ELEC 431 or ELEC 531 - for Fourier transforms and discrete-time systems

ELEC 430 or ELEC 533 - for probability and random processes

Experience with Matlab (helpful)

Grading

35% - Test

35% - Hands-on group project

20% - Homework

10% - Notebook and classroom participation

Study Groups
To encourage group learning, students are expected to form study groups. Homeworks may be completed in groups. On a weekly basis, each group must turn in (with homework solutions) a report on their activities and a summary of the material covered in the previous week of class. Group work should not substitute for study on your own. At test time, only a pencil will accompany you.

Late Homework Policy
Homework is due before midnight on the due date. Please slip your homeworks under Justin Romberg's door (DH 2142). After the due date, but before solutions are handed out, homework can be turned in for 50% credit. After solutions are handed out, 0% credit will be issued. However, all assignments must be turned in, or an incomplete grade will be assigned. You are encouraged to work in groups on homework problems, as long as you ultimately formulate your own solution. You are expected to understand any solution you turn in.
Homework, tests, and solutions from previous offerings of this course are off limits, under the honor code.

Handouts Policy
Course notes and assignment solutions will be handed out in class. Additional handouts will not be provided. Each student is responsible for obtaining notes for classes not attended.

Suggestions

Remember the big picture

Read the book and supplementary sources

Prepare your own summaries from texts and notes

Work in groups for homeworks and study (explain main concepts to each other)

COURSE OUTLINE

I. Introduction

Motivation: Why spectral analysis?

Review of linear algebra

Review of random processes

II. Classical Spectrum Estimation

Periodogram

Windowing and averaging

III. Parametric Modeling

Linear prediction

Levinson algorithm

AR spectrum estimation

MA and ARMA spectrum estimation

IV. Advanced Techniques

Minimum variance method

Eigenspace methods (MUSIC, etc.)

V. Thomson's Multiple Window Method

Theoretical background

Optimal windows

Line component estimation

VI. Applications

Time-varying spectral analysis

Wavelets

Speech processing

Tue Jan 18 2000.