2004 IEEE International Conference on Acoustics, Speech, and Signal Processing
DOI: 10.1109/icassp.2004.1326206
|View full text |Cite
|
Sign up to set email alerts
|

Range and bearing estimation of wideband sources using an orthogonal beamspace processing structure

Abstract: We propose a new method for range and bearing estimation of wideband sources based on a novel beamspace structure. The proposed structure consists of an orthogonal set of beamspace processors that have a frequency-invariant beampattern property, and can be steered easily to any nearfield range while maintaining the frequency-invariance of the spatial response. These properties mean that the proposed structure can perform coherent signal subspace processing for nearfield wideband sources, making it ideal for wi… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
10
0

Publication Types

Select...
3
3

Relationship

0
6

Authors

Journals

citations
Cited by 13 publications
(10 citation statements)
references
References 13 publications
0
10
0
Order By: Relevance
“…In comparison with the other aforementioned approaches, 21 only one generalized eigenvalue decomposition is necessary, thus limiting the computation cost of the method. Its implementation in a coherent beamspace paradigm is proposed along the lines of [72] and an original constructive method is proposed to the synthesis of focalization matrices, in a convex optimization setup. Besides, the approach is able to deal with reverberant robotics environments, since the statistical independence of the sources together with their mutual independence w.r.t.…”
Section: Musicmentioning
confidence: 99%
“…In comparison with the other aforementioned approaches, 21 only one generalized eigenvalue decomposition is necessary, thus limiting the computation cost of the method. Its implementation in a coherent beamspace paradigm is proposed along the lines of [72] and an original constructive method is proposed to the synthesis of focalization matrices, in a convex optimization setup. Besides, the approach is able to deal with reverberant robotics environments, since the statistical independence of the sources together with their mutual independence w.r.t.…”
Section: Musicmentioning
confidence: 99%
“…A number W = 16 of FFT results, computed over successive disjoint 1024-point rectangular temporal windows, is then collected to estimate the covariance matrix C X . In all the following, only the frequencies within [300Hz; 3kHz] are used for localization, so that about B = 180 frequency bins are considered in order to form the MUSIC pseudo-spectrum (10).…”
Section: Localization Resultsmentioning
confidence: 99%
“…. , B, as was done in Algorithm 1; -from the spectral decomposition ofĈ X (k b ), compute the projectorΠ N (k b ) and the pseudo-spectrum h b (r, θ ) related to the b th bin; end -isolate the sources locations as the maxima of the pseudo-spectrum h naive (r, θ ) defined in (10). end Algorithm 2: MUSIC naive extension to broadband …”
Section: B a Naive Extensionmentioning
confidence: 99%
“…From the phase values recovered at each element, it is possible to compute the position of the source. There have been numerous studies on algorithms to compute near-field angles of arrival [7][8][9][10][11][12][13], although for this study, we have chosen to apply a relatively simple triangulation method, similar to that used in [14]. This gives a better appreciation of the result that can be obtained using a straightforward system, with minimal processing overhead, that is, to produce an instantaneous result.…”
Section: Near-field Position Detectionmentioning
confidence: 99%