Robust adaptive beamforming based on a new steering vector
estimation algorithm
Weimin Jia
a,
n
, Wei Jin
a
, Shuhua Zhou
b
, Minli Yao
a
a
Department of Electronic and Information Engineering, Xi'an Research Institute of High Technology, Xi'an, Shaanxi 710025, China
b
Wire Communication Technique Office, Xi'an Research Institute of High Technology, Qing Zhou, Shandong 262500, China
article info
Article history:
Received 16 December 2012
Received in revised form
6 February 2013
Accepted 11 March 2013
Available online 18 March 2013
Keywords:
Adaptive beamforming
Steering vector estimation
Quadratically constrained quadratic
programming
Robustness
abstract
A new approach to the design of robust adaptive beamforming is introduced. In the
proposed approach, the mismatch vector of the desired steering vector is estimated
by solving a quadratically constrained quadratic programming problem using an
interference-plus-noise subspace projection matrix. The presumed look direction of
desired signal is the only prior informa tion of the proposed approach, and the parameters
of uncertainty set or the angular sectors of the desired signal are not needed. In the
presence of large DOA mismatch, the proposed beamfor mer performs well. Moreover, the
proposed approach can deal with arbitrary steering vector mismatch in theor y while
many existing advanced robust beamformers cannot. Hence, it is very suitable for many
practical applications.
Crown Copyright & 2013 Published by Elsevier B.V. All rights reserved.
1. Introduction
Adaptive beamforming has been widely used in wire-
less communications, microphone array speech proces-
sing, radar, sonar, medical imaging, radio astronomy, and
other areas [1]. It has high performance in interference
suppression if the array steering vector corresponding to
the desired signal is accurately known. However, the
adaptive beamformer is also well-known to be sensitive
to model mismatch, especially when the desired signal is
present in the training data. Therefore, many approaches
have been proposed to improve the robustness of the
adaptive beamformer.
The most popular conventional robust adaptive beam-
formers are diagonal loading beamformer [2] and
eigenspace-based beamformer [3]. However, the main
shortcoming of the former approach is that it is not clear
how to obtain the optimal value of the diagonal load-
ing factor, whereas the latter approach is essentially
ineffective at low signal-plus-interference ratios (SNRs)
and when the dimension of the signal-plus-interference
subspace is high. In [4], a modified projection approach is
proposed and it can work well even at low SNRs, but it
does not work well in the case of large steering vector
mismatch.
Recently, based on worst-case performance optimiza-
tion and uncertainty set of the desired array steering
vector, the robust beamformers which can combat steering
vector mismatch effectively are proposed [5,6]. In addition,
a direction-of-arrival (DOA) uncertainty based beamfor-
mer that is robust against DOA mismatch is proposed [7].
Very recently, some advanced robust beamforming
approaches using as little as possible prior information
have been proposed [8–11]. One common ground of these
approaches is that they all use an integration matrix
related to the array steering vector within some special
angular sectors. This implies that in these approaches, the
array antenna is assumed without any calibration error.
Though these methods are effective on many mismatches,
such as look direction error, signal spatial signature mis-
match due to coherent local scattering or incoherent local
Contents lists available at SciVerse ScienceDirect
journal home page: www.elsevier.com/locate/sigpro
Signal Processing
0165-1684/$ - see front matter Crown Copyright & 2013 Published by Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.sigpro.2013.03.015
n
Corresponding author.
E-mail addresses: jwm907@163.com (W. Jia),
jinweimail@126.com (W. Jin), zsh0214@yahoo.com.cn (S. Zhou),
yaominli@sohu.com (M. Yao).
Signal Processing 93 (2013) 2539 – 2542