Approaches to the Processing of Data from Large Aperture Acoustic Vertical Line Arrays
Abstract
This dissertation investigates various approaches to the processing of narrowband frequency data from large aperture acoustic vertical line arrays with on the order of one hundred equally spaced sensors. When little is known about the oceanic environment, beamforming is performed to detect signals and estimate the vertical arrival structure of the pressure field. The attention is focussed on using the adaptive Minimum Variance processor. The issues of signal cancellation due to correlated arrivals and mismatch due to wavefront curvature are examined. In practice, planar wavefront replica vectors do not result in significant mismatch for a 900 m aperture. Processing subapertures is proposed because of the adaptive processor requirement of having numerous data snapshots to estimate and invert the covariance matrix, while the pressure field may not be stationary. Full aperture and subaperture processing techniques are validated with simulated data created by a normal mode model, and then used on 200 Hz data collected in the North-East Pacific by the Marine Physical Laboratory in September 1987. Multipath arrivals are found to illuminate only parts of the 120 sensor array, cut for 100 Hz and 900 m long, thus indicating that the wavefield can be highly inhomogeneous with depth. (r.h.)
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 01, 1990
- Accession Number
- ADA225297
Entities
People
- Jean-marie Q. Tran
Organizations
- Scripps Institution of Oceanography