Cintal logo

Vector Sensor Array Processing and Geoacoustic Inversion with MAKAI'05 Sea Trial Data

B. Jiang,
SiPLAB, University of Algarve, Campus de Gambelas,
PT-8005-139 Faro, Portugal

Comments: download file (pdf)
Ref.: SiPLAB Report 03/13, SiPLAB, University of Algarve,2011.

Underwater acoustic vector sensor has received much more attentions in recent a few decades. One type of high precise high-frequency vector sensor TV-001, succeeded for several applications in Makai Ex05 sea trial, such as direction-of-arrival (DOA) estimation, geoacoustic inversion, tomography, MIMO communications. In this report, the vector sensor array (VSA) direction finding performance, including the array directivity index (DI), array gain (AG), and Cramér-Rao bound (CRB) are studied. Besides, the main research was focused on the geoacoustic inversion by using various propagation models. One simple inversion method by comparing the downward and upward beams was studied, i.e. bottom reflection coefficient (BRC) matching method, the results showed that the (p+v) processor is the best. The seabed parameter sensitivities were studied, and the compressional velocity was found could be inverted by matched-field inversion (MFI) methods. The CRB's of geoacoustic parameter estimates are derived, which also demonstrated that the (p+v) processing outperforms others. A two-step inversion method was proposed. First, the compressional velocity, along with the receiver range and depth were optimized using genetic algorithm (GA). Secondly, the optimized range and depth information was fed back to improve the accuracy of the replica fields, then the compressional velocity dependent replicas were matched with the real data, giving high resolution and precise results during the period of nearly two hours.

ACKNOWLEDGMENT: this work was supported by project SENSOCEAN funded under the Foundation for Science and Technology (Portugal).