Evaluation of Robust Capon Beamforming for Passive Sonar

Somasundaram, S.; Parsons, Nigel H.

doi:10.1109/joe.2011.2167374

Cited by 22 publications

(28 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We now describe the approach outlined in [16], [17] for designing a spherical ASV uncertainty set for the bth beam and for the frequency bin with center frequency f , assuming azimuth and elevation angle steering, so that θ = [θ, φ]…”

Section: B Uncertainty Set Designmentioning

confidence: 99%

“…The main benefit of the RCB-type techniques, which are the focus of this work, is that their constraint is the ellipsoidal uncertainty set of the array steering vector, which can be defined in a systematic manner to allow for mismatch in passive sonar applications [16]- [19]. We remark that the benefits of using a narrowband robust Capon beamformer (NBRCB) over the commonly used white noise gain constrained MVDR beamformer, with or without extra linear constraints, were shown in [16] for passive sonar.…”

Section: Introductionmentioning

confidence: 99%

“…However, as was shown in [12], the RCB weights are simpler to compute and the RCB method provides a means for addressing the scaling ambiguity issue occurring in WC-RABs. For this reason, several of our previous works have focused on RCB based beamforming [16]- [19].…”

Section: Introductionmentioning

confidence: 99%

“…Initially, we compare and contrast these efficient algorithms using simulated passive sonar data and then select the most promising algorithms to implement the RCB-based NBRCB algorithm used in [16], [19]. These efficient implementations are evaluated on in-water recorded experimental data and compared with the original RCB-EVD based implementation.…”

Section: Introductionmentioning

confidence: 99%

“…The uncertainty sets must be specified for each beam and FFT bin (see, e.g., [2], [14], [16], [17], [19], [41] for ellipsoidal uncertainty set design techniques). Denoting the solution to (6)â 0,θ b ,f (see also [2], [11], [12]), to overcome the scaling ambiguity, the ASV is scaled so that its Euclidean length squared equals M , forminĝ…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Low-Complexity Uncertainty-Set-Based Robust Adaptive Beamforming for Passive Sonar

Somasundaram

Butt

Jakobsson

et al. 2024

IEEE J. Oceanic Eng.

Self Cite

View full text Add to dashboard Cite

Recent work has highlighted the potential benefits of exploiting ellipsoidal uncertainty set based robust Capon beamformer (RCB) techniques in passive sonar. Regrettably, the computational complexity of forming RCB weights is cubic in the number of adaptive degrees of freedom, which is often prohibitive in practice. For this reason, several low-complexity techniques for computing RCB weights, or equivalent worst-case robust adaptive beamformer (WC-RAB) weights, have recently been developed, whose complexities are only quadratic in the number of adaptive degrees of freedom. In this work, we review several such techniques for passive sonar, evaluating them initially on simulated data. The best performing methods are then evaluated on in-water recorded passive sonar data. Index TermsEfficient robust adaptive beamforming, robust Capon beamforming, worst-case robust adaptive beamforming, robust adaptive beamforming, underwater acoustics. I. INTRODUCTIONIn this work, we are interested in examining techniques for efficiently computing ellipsoidal uncertainty set based robust Capon beamformer (RCB) weights, or equivalent worst-case robust adaptive beamformer (WC-RAB) weights, for the purposes of implementing recent RCB-based passive sonar beamformers with low computational complexity.As is well known, an adaptive beamformer can outperform a conventional delay-and-sum beamformer providing that it a) is robust to array steering vector errors, b) can converge sufficiently on the, often limited, data available, and c) is low enough complexity to be implementable on the, typically limited, computational resources that are available [1], [2]. The model of the signal-of-interest spatial signature, or array steering vector, that is used by a beamformer is usually derived under plane-wave assumptions, as a set of complex phase delays, and is subject to several sources of error. These include calibration errors, due to sensor gain/phase errors, mutual coupling effects and sensor position errors, pointing errors, caused by differences in the assumed angle-of-arrival and the true one, and deviations from plane wave assumptions, e.g., caused by inhomogeneity in the ocean and/or multipath propagation. If not properly dealt with, these errors lead to cancellation of the desired signal and a consequent loss of array output signal to noise ratio [3] [14]. The main practical drawback with many robust adaptive algorithms is that the user is a required to select a parameter that is not directly related to the steering vector uncertainty, making the parameter choice difficult and often ad hoc. For instance, it is unclear how to choose a white noise gain constraint, which is a constraint on the norm of the weights, based on the array steering vector uncertainty [15]. The main benefit of the

show abstract

Section: B Uncertainty Set Designmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%