Course Number:

Credit Hours:

4

Course Title:

Modern Spectral Analysis and Array Processing

Course Description:

Describes the problem of estimating spectra from finite records of noisy data and reviews applications including communications (especially wireless), biomedicine, geophysics, speech, nondestructive testing, and sonar and radar. Explores common power spectrum estimation algorithms. Emphasizes the advantages and limitations of conventional, Capon’s, multiple window, maximum entropy, parametric (AR, MA, and ARMA), and harmonic decomposition (Prony, Pisarenko, and SVD) methods, in terms of accuracy (bias), reliability (variance), applicability, and other criteria. Introduces higher-order and nonstationary spectrum estimation including conventional and parametric higher-order methods and sliding window (short-time Fourier transform and model-based), adaptive, time-frequency, and wavelet techniques for the nonstationary problem. Examines extensions to multichannel and multidimensional data, discusses the array processing problem from a spectrum estimation perspective, and introduces the wave-field perspective. Discusses nonparametric and parametric array processing techniques and applications.