Fast communication
Off-grid DOA estimation using array covariance matrix and
block-sparse Bayesian learning
$
Yi Zhang
a,b
, Zhongfu Ye
a,b,
n
,XuXu
a,b
, Nan Hu
a,b
a
Department of Electronic Engineering and Information Science, University of Science and Technology of China,
Hefei, Anhui 230027, China
b
National Engineering Laboratory for Speech and Language Information Processing, Hefei, Anhui 230027, China
article info
Article history:
Received 30 July 2013
Received in revised form
25 October 2013
Accepted 20 November 2013
Available online 28 November 2013
Keywords:
Direction-of-arrival
Off-grid model
Sparse recovery
Sparse Bayesian learning
abstract
A new method based on a novel model for off-grid direction-of-arrival (DOA) estimation is
presented. The novel model is based on the sample covariance matrix and the off-grid
representation of the steering vector. Based on this model, its equivalent signals are
assumed to satisfy independent Gaussian distribution and its noise variance can be
normalized to 1. The off-grid DOAs are estimated by the block sparse Bayesian algorithm.
The advantages of the proposed method are that it considers the temporal correlation
existed in each row of the equivalent signal sample matrix and the normalized noise
variance does not need to be estimated. Moreover, this algorithm can work without the
knowledge of the number of signals. Numerical simulations demonstrate the superior
performance of the proposed method.
& 2013 Elsevier B.V. All rights reserved.
1. Introduction
Direction-of-arri v al (DOA) estimation using sensor arrays
plays a fundamental role in various applications such as
acoustic source localization, radar imaging, mobile commu-
nication, wireless sensor networks, etc. [1,2]. Many subspace-
based approaches [3,4] hav e been proposed.
In recent years, by exploiting the spatial sparsity in the
array model, some sparsity-driven methods have been
presented. The most successful one is ℓ
1
SVD [5] which
employed ℓ
1
norm to enforce sparsity and singular value
decomposition (SVD) to reduce complexity and sensitivity
against noise. The sparse recovery for weighted subspace
fitting (SRWSF) [6] improved the DOA estimation accuracy
of ℓ
1
SVD by using weighted subspace fitting. Both these
sparse recovery based methods employ fixed sampling
grid, and assume that all the true DOAs are exactly located
on the selected grid. When the true DOAs are beyond the
fixed grid, their performance will degrade due to errors
caused by mismatches. Zhu et al. [7] introduced the
off-grid model for DOA estimation and proposed a sparse
total least squares (STLS) method based on the Gaussian
assumption of off-grid distance, which, however, is not
satisfied in the off-grid DOA estimation problem. In [8], the
off-grid distance is assumed to satisfy uniform prior and a
new off-grid algorithm termed OGSBI-SVD was formulated
from a Bayesian perspective. The Bayesian inference was
employed to estimate the off-grid DOAs. However, the
Bayesian inference method suffered from the underesti-
mation of the noise variance.
In this communication, a novel model based on the
sample covariance matrix and the off-grid representation
of the steering vector is firstly derived. According to the
statistical property of the sample covariance matrix, the
noise variance of the novel model is normalized to 1. All
the equivalent signals of the new model are assumed to
satisfy the independent Gaussian distribution and the
Contents lists available at ScienceDirect
journal home page: www.elsevier.com/locate/sigpro
Signal Processing
0165-1684/$ - see front matter & 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.sigpro.2013.11.022
☆
This work is supported by the National Natural Science Foundation of
China (no. 61101236).
n
Corresponding author at: Department of Electronic Engineering and
Information Science, University of Science and Technology of China,
Hefei, Anhui 230027, China. Tel.: þ86 551 63601314.
E-mail address: yezf@ustc.edu.cn (Z. Ye).
Signal Processing 98 (2014) 197–201