Course Description
Processing techniques for continuous and discrete signals with initially unknown or time-varying characteristics. Parameter estimation; Bayes, MAP, maximum likelihood, least squares the Cramer-Rao bound. Linear estmation, prediction, pwer spectrum estmation, lattice filters. Adaptive filtering by LMS and recursive least squares. Kalman filtering. Eigenmethods for spectral estimation. Implementation issues and numerical methods of computation are considered throughout. Prerequisite: ENSC 802 and 429 or their equivalents.
Textbooks
S. Haykin, 'Adaptive Filter Theory, 3rd ed.', Prentice-Hall, 1996.
Reference Books
Louis Scharf, 'Statistical Signal Processing', Addison-Wesley, 1991.
Prerequisites
Successful completion of <a href=/grad/courses/ENSC802.html>ENSC 802-3</a> and a previous course in DSP at the undergraduate level is required for students wishing to take this course.