PH4454 Sonar Signal and Array Processing

A treatment of the fundamental phenomena basic to sound propagation in the ocean and its application to sonar signal and array processing. Topics include individual terms of the sonar equation for single and multiple hydrophone sensors, and signal processing of these data for a sonar display. Particular subjects include units, such as Power and Power Spectral Density in conventional and adaptive beamforming applications for source parameter estimation. Additional topics covered are the discrete Fourier transform, effects of temporal and spatial shading, auto-, cross-correlation, and convolution in the presence of ideal and observed ocean noise. The student will execute and demonstrate learning objectives via a programming language in simulations and with data.

Prerequisite

PH3452 (may be taken concurrently)

Lecture Hours

4

Lab Hours

2

Course Learning Outcomes

Upon successful completion of this course, a student will:

• Apply discrete Fourier transform (DFT) to interpret acoustic simulations and recorded data
• Compute auto- and cross-correlation and convolution with data and investigate how their computations differ to recover a channel impulse response or implement matched-filtering approaches with single sensors
• Estimate conventional power spectra by calculating and converting levels between Power and Power Spectral Density with temporal and spatial shading for a sinusoid received at an array of sensors using the decibel scale
• Apply beamforming principles and implement plane-wave phase delays in the time- and frequency-domain for a linearly spaced array of N sensors
• Calculate conventional weights and implement a conventional beamforming (CBF) algorithm in the time- and frequency-domain
• Investigate the spatial covariance matrix structure and its properties with applications to CBF
• Calculate adaptive weights and implement an adaptive beamforming (ABF) algorithm (MVDR and WNGC) in the frequency domain to localize one or multiple sources
• Know the principles and requirements for ABF (uncorrelated sources, full rank and invertible sample covariance matrix)
• Implement and evaluate additive white Gaussian noise defined for single-sensor or array SNR in the frequency domain
• Execute and demonstrate learning objectives via a programming language (e.g., MATLAB) in simulations and with data
• Be able to summarize scientific papers in assignments