Multisnapshot Sparse Bayesian Learning for DOA

Peter Gerstoft; Christoph F. Mecklenbrauker; Angeliki Xenaki; Santosh Nannuru

doi:10.1109/LSP.2016.2598550

Multisnapshot Sparse Bayesian Learning for DOA

Peter Gerstoft, Christoph F. Mecklenbrauker, Angeliki Xenaki, Santosh Nannuru

Research output: Contribution to journal › Journal article › Research › peer-review

Abstract

The directions of arrival (DOA) of plane waves are estimated from multisnapshot sensor array data using sparse Bayesian learning (SBL). The prior for the source amplitudes is assumed independent zero-mean complex Gaussian distributed with hyperparameters, the unknown variances (i.e., the source powers). For a complex Gaussian likelihood with hyperparameter, the unknown noise variance, the corresponding Gaussian posterior distribution is derived. The hyperparameters are automatically selected by maximizing the evidence and promoting sparse DOA estimates. The SBL scheme for DOA estimation is discussed and evaluated competitively against LASSO (l(1)-regularization), conventional beamforming, and MUSIC.

Original language	English
Journal	IEEE Signal Processing Letters
Volume	23
Issue number	10
Pages (from-to)	1469-1473
Number of pages	5
ISSN	1070-9908
DOIs	https://doi.org/10.1109/LSP.2016.2598550
Publication status	Published - 2016

Keywords

Array processing
compressive beamforming
directions of arrival (DOA) estimation
relevance vector machine
sparse reconstruction
Beamforming
Gaussian distribution
Complex Gaussian
Conventional beamforming
Posterior distributions
Relevance Vector Machine
Sparse Bayesian learning
Sparse Bayesian learning (SBL)
Sparse reconstruction
Zero-mean complex
Direction of arrival
Direction-of-arrival estimation
Covariance matrices
Arrays
Bayes methods
Estimation
Array signal processing
Signal processing algorithms
Arrayprocessing
compressivebeamforming
Signal processing and detection
Sensing devices and transducers
Other topics in statistics
Digital signal processing
Knowledge engineering techniques
direction-of-arrival estimation
learning (artificial intelligence)
sensor arrays
sparse Bayesian learning
directions of arrival
plane waves
multisnapshot sensor array data
SBL
source amplitudes
independent zero-mean complex Gaussian distribution
hyperparameters
source powers
complex Gaussian likelihood
DOA estimation

Access to Document

10.1109/LSP.2016.2598550

https://arxiv.org/pdf/1602.09120.pdf

OpenUrl availability

Full text

Cite this

@article{fd33bfff14114b829bfe443983005a19,

title = "Multisnapshot Sparse Bayesian Learning for DOA",

abstract = "The directions of arrival (DOA) of plane waves are estimated from multisnapshot sensor array data using sparse Bayesian learning (SBL). The prior for the source amplitudes is assumed independent zero-mean complex Gaussian distributed with hyperparameters, the unknown variances (i.e., the source powers). For a complex Gaussian likelihood with hyperparameter, the unknown noise variance, the corresponding Gaussian posterior distribution is derived. The hyperparameters are automatically selected by maximizing the evidence and promoting sparse DOA estimates. The SBL scheme for DOA estimation is discussed and evaluated competitively against LASSO (l(1)-regularization), conventional beamforming, and MUSIC.",

keywords = "Array processing, compressive beamforming, directions of arrival (DOA) estimation, relevance vector machine, sparse reconstruction, Beamforming, Gaussian distribution, Complex Gaussian, Conventional beamforming, Posterior distributions, Relevance Vector Machine, Sparse Bayesian learning, Sparse Bayesian learning (SBL), Sparse reconstruction, Zero-mean complex, Direction of arrival, Direction-of-arrival estimation, Covariance matrices, Arrays, Bayes methods, Estimation, Array signal processing, Signal processing algorithms, Arrayprocessing, compressivebeamforming, Signal processing and detection, Sensing devices and transducers, Other topics in statistics, Digital signal processing, Knowledge engineering techniques, direction-of-arrival estimation, learning (artificial intelligence), sensor arrays, sparse Bayesian learning, directions of arrival, plane waves, multisnapshot sensor array data, SBL, source amplitudes, independent zero-mean complex Gaussian distribution, hyperparameters, source powers, complex Gaussian likelihood, DOA estimation",

author = "Peter Gerstoft and Mecklenbrauker, {Christoph F.} and Angeliki Xenaki and Santosh Nannuru",

year = "2016",

doi = "10.1109/LSP.2016.2598550",

language = "English",

volume = "23",

pages = "1469--1473",

journal = "IEEE Signal Processing Letters",

issn = "1070-9908",

publisher = "Institute of Electrical and Electronics Engineers",

number = "10",

}

TY - JOUR

T1 - Multisnapshot Sparse Bayesian Learning for DOA

AU - Gerstoft, Peter

AU - Mecklenbrauker, Christoph F.

AU - Xenaki, Angeliki

AU - Nannuru, Santosh

PY - 2016

Y1 - 2016

N2 - The directions of arrival (DOA) of plane waves are estimated from multisnapshot sensor array data using sparse Bayesian learning (SBL). The prior for the source amplitudes is assumed independent zero-mean complex Gaussian distributed with hyperparameters, the unknown variances (i.e., the source powers). For a complex Gaussian likelihood with hyperparameter, the unknown noise variance, the corresponding Gaussian posterior distribution is derived. The hyperparameters are automatically selected by maximizing the evidence and promoting sparse DOA estimates. The SBL scheme for DOA estimation is discussed and evaluated competitively against LASSO (l(1)-regularization), conventional beamforming, and MUSIC.

AB - The directions of arrival (DOA) of plane waves are estimated from multisnapshot sensor array data using sparse Bayesian learning (SBL). The prior for the source amplitudes is assumed independent zero-mean complex Gaussian distributed with hyperparameters, the unknown variances (i.e., the source powers). For a complex Gaussian likelihood with hyperparameter, the unknown noise variance, the corresponding Gaussian posterior distribution is derived. The hyperparameters are automatically selected by maximizing the evidence and promoting sparse DOA estimates. The SBL scheme for DOA estimation is discussed and evaluated competitively against LASSO (l(1)-regularization), conventional beamforming, and MUSIC.

KW - Array processing

KW - compressive beamforming

KW - directions of arrival (DOA) estimation

KW - relevance vector machine

KW - sparse reconstruction

KW - Beamforming

KW - Gaussian distribution

KW - Complex Gaussian

KW - Conventional beamforming

KW - Posterior distributions

KW - Relevance Vector Machine

KW - Sparse Bayesian learning

KW - Sparse Bayesian learning (SBL)

KW - Sparse reconstruction

KW - Zero-mean complex

KW - Direction of arrival

KW - Direction-of-arrival estimation

KW - Covariance matrices

KW - Arrays

KW - Bayes methods

KW - Estimation

KW - Array signal processing

KW - Signal processing algorithms

KW - Arrayprocessing

KW - compressivebeamforming

KW - Signal processing and detection

KW - Sensing devices and transducers

KW - Other topics in statistics

KW - Digital signal processing

KW - Knowledge engineering techniques

KW - direction-of-arrival estimation

KW - learning (artificial intelligence)

KW - sensor arrays

KW - sparse Bayesian learning

KW - directions of arrival

KW - plane waves

KW - multisnapshot sensor array data

KW - SBL

KW - source amplitudes

KW - independent zero-mean complex Gaussian distribution

KW - hyperparameters

KW - source powers

KW - complex Gaussian likelihood

KW - DOA estimation

U2 - 10.1109/LSP.2016.2598550

DO - 10.1109/LSP.2016.2598550

M3 - Journal article

SN - 1070-9908

VL - 23

SP - 1469

EP - 1473

JO - IEEE Signal Processing Letters

JF - IEEE Signal Processing Letters

IS - 10

ER -

Multisnapshot Sparse Bayesian Learning for DOA

Abstract

Keywords

Access to Document

OpenUrl availability

Fingerprint

Cite this