Abstract
New quasi-Bayesian (hybrid) Cramér-Rao bound (CRB) expressions are herein derived for far-field deep-sea direction-of-arrival (DOA) estimation with a nominally linear towed-array that 1) is deformed by spatio-temporally correlated oceanic currents, which have been previously overlooked in the towed-array shape-deformation statistical analysis literature, 2) is deformed by temporally correlated motion of the towing vessel, which is modeled only as temporally uncorrelated in prior literature, and 3) suffers gain-uncertainties and phase-uncertainties in its constituent hydrophones. This paper attempts to bridge an existing literature gap in deformed towed-array DOA-estimation performance analysis, by simultaneously a) incorporating several essential fluid-mechanics considerations to produce a shape-deformation statistical model physically more realistic than those previously used for DOA performance analysis and b) rigorously derive a mathematical analysis to characterize quantitatively and qualitatively the DOA stimation's statistical performance. The derived CRB expressions are parameterized in terms of the towed-array's physically measurable nonidealities for the single-source case. The new hybrid-CRB expressions herein derived are numerically more stable than those in the current literature.
Original language | English |
---|---|
Pages (from-to) | 36-47 |
Number of pages | 12 |
Journal | IEEE Transactions on Signal Processing |
Volume | 52 |
Issue number | 1 |
DOIs | |
Publication status | Published - 1 Jan 2004 |
Externally published | Yes |
Keywords
- Acoustical signal processing
- Array signal processing
- Direction-of-arrival estimation
- Marine telemetry
- Parameter estimation
- Sonar arrays
- Sonar signal processing
- Underwater acoustic arrays
ASJC Scopus subject areas
- Signal Processing
- Electrical and Electronic Engineering