This course focuses on the problem of determining the spectral distribution of a time series (random process) from a finite set of measurements. Spectral estimation techniques can applied to various sensors: active and passive radar in addition to sonar. They can be used in localization of targets, imaging and its interpretation, and direction of arrival (DOA) estimation, etc. Students should learn various spectral estimation paradigms such as nonparametric methods based on FFT, parametric methods for rational (continuous) spectra, parametric methods for line (discrete) spectra, and filter bank methods. Special emphases will be paid upon the application of spectral estimation methods to various kinds of radar signal processing.

Signals and Systems, Discrete-Time Signal Processing (DSP), Random Process, Linear Algebra

Mid-Exam(30%), Homeworks(30%), Term Project(40%)
Note: There will be no points for copied homeworks.

1) Statistical and Adaptive Signal Processing, Dimitris G. Manolakis, Vinay K. Ingle, and Stephen
M. Kogon, McGraw Hill 2000
2) Digital spectral analysis with applications, S. L. Marple Jr., Prentice Hall, 1987
3)Modern spectral estimation, S. M. Kay, Prentice Hall, 1988

· Random Process and Fourier Transform - Review
· Basic Concepts
· Nonparametric methods
· Mid-Term Exam
· Parametric methods for rational spectra
· Parametric methods for line spectra
· Filter bank methods
· Term Project Presentation