H. Gazzah, houssem.gazzah@isen-ouest.yncrea.fr
ISEN Yncrea Ouest, L@bisen/SEACOM,
44470 Carquefou, France
S.M. Jesus sjesus@ualg.pt
LARSys, University of Algarve,
Campus de Gambelas, PT-8005-139 Faro, Portugal.
Comments: download (pdf).
Ref.: Forum Acousticum, Torino (Italy), September 2023.
Abstract
High-resolution estimation of modal functions has been recently demonstrated. Two VLAs collect pressure data
due to a monochromatic source, and feed a subspace algorithm that computes, in a fully automatic closed-form
manner, modal functions at sensors depths. Estimation accuracy can be improved at will by improving the SNR,
using the same limited number of sensors. In this proposal, the apparatus is improved in order to reduce by
two the number of required sensors to cover a given portion of the water column, or equivalently, to double
the covered portion of the water column using the same number of sensors. This is made possible by allowing
the source and the hydrophones to be deployed at different depth locations. The modified algorithm continues
to exhibit the same attractive features of the original one, i.e. closed-form and asymptotically (as SNR
increases) unbiased estimation. At limited SNR, a moderate degradation of the estimates is observed because
the collected data matrices are not symmetric anymore.