Masters Theses

Date of Award

5-1995

Degree Type

Thesis

Degree Name

Master of Science

Major

Electrical Engineering

Major Professor

Michael J. Roberts

Committee Members

Daniel B. Koch, Robert Bodenheimer

Abstract

An algorithm to reconstruct a composite time series output from the data generated by a multiple-aperture frequency-domain beamformer was conceived, designed, and implemented for an underwater acoustic measurement system. Each aperture of the acoustic measurement system has multiple analog sensor inputs. The sample rate is different for each aperture, and each aperture is optimized for measurements over a different frequency range. The beamformer produces a set of frequency-domain spectral estimate measurement data for each aperture that is used as the input to the composite time series reconstruction algorithm. The signal processing operations required to reconstruct the composite time series output are identified as the inverse fast Fourier Transform (FFT), upsampling, and spectral merging. The inverse FFT is used to generate an individual time series output for each aperture. Since the apertures have different sample rates and overlapping frequency ranges, the individual aperture time series are summed via successive application of the upsampling and spectral merging operations. The design rationale for the digital filters used in the upsampling and spectral merging operations is presented in detail. Expressions describing the relationship in absolute time between the individual aperture time series for adjacent apertures are derived and used to design the data buffering scheme required to produce a continuous composite time series output. Constraints on the available execution time within the beamformer require simplifications in the beamforming operation and the composite time series reconstruction algorithm. The impact of these simplifications on the accuracy of the composite time series output is discussed in some detail. The major impact on the accuracy of the composite time series reconstruction algorithm is identified as ± 0.09 dB of ripple introduced by simplifications to the upsampling and spectral merging digital filters. Several types of tests were used to verify both the composite time series reconstruction algorithm and its implementation. A brief discussion of each type of test is presented, along with representative results from each type of test. The results of end-to-end system testing with real analog inputs reveal additional inaccuracies introduced into the composite time series output by the acoustic measurement system analog signal conditioning electronics. As a result, a recommendation is made to determine a frequency-dependent calibration correction for the spectral estimate input data. The correction is necessary before the composite time series output can be used as an absolute measure of underwater sound.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS