Contents lists available at ScienceDirect
Signal Processing
journal homepage: www.elsevier.com/locate/sigpro
Short communication
Angle estimation for bistatic MIMO radar in the presence of spatial colored
noise
☆
Fangqing Wen
a,
⁎
, Xiaodong Xiong
a
, Jian Su
b
, Zijing Zhang
c
a
Electronic and Information School of Yangtze University, Jingzhou 434023, China
b
School of Communication and Information Engineering, University of Electronic Science and Technology of China, Chengdu 611731, China
c
National Laboratory of Radar Signal Processing, Xidian University, Xi'an 710071, China
ARTICLE INFO
Keywords:
Colocated MIMO radar
Bistatic MIMO radar
Angle estimation
Spatial colored noise
ABSTRACT
In this paper, we address the problem of angle estimation for bistatic multiple-input multiple-output (MIMO)
radar in the presence of spatial colored noise. By exploiting the temporal structure and the multidimensional
inherent structure, a fourth-order cross-covariance tensor is formulated to eliminate the effect of spatial colored
noise. The higher-order singular value decomposition is utilized for accurate signal subspace estimation.
Thereafter, angles are obtained with the shift-invariance technique. The proposed method can achieve
automatic pairing of the estimated angles without any virtual aperture loss, thus it has better estimation
performance than existing methods. Extensive numerical experiments verify the effectiveness and improvement
of our algorithm.
1. Introduction
Accompanied by the demanding requirements for location accura-
cies and resolutions in a radar system, multiple-input multiple-output
(MIMO) radar has aroused extensive attention in the past decade [1].
MIMO radar system simultaneously transmits mutually orthogonal
waveforms with multiple antennas and receives the reflected echoes
with multiple antennas. Theoretical research indicates that MIMO
radar has several built-in advantages in noise suppression, overcoming
the fading effect, improving spatial resolution, enhancing parameter
identifiability, etc [2–5]. Generally speaking, MIMO radar can be
divided into two classes according to the antenna architectures. The
first kind is statistical MIMO radar [6], which takes advantage of widely
separated antennas to solve target scintillation problem. The other kind
is colocated MIMO radar [7], including monostatic MIMO radar and
bistatic MIMO radar [8], whose antennas are closely spaced in both
transmit array and receive array to obtain unambiguous angle estima-
tion. The array antenna offers great flexibility for system design in
combination with suitable control and signal processing. In this paper,
we stress the problem of direction-of-departure (DOD) and direction-
of-arrival (DOA) estimation in bistatic MIMO radar.
Angle estimation is a canonical problem in colocated MIMO radar
that has become a hot research topic. Various methods have been
proposed, such as Capon [9], MUSIC [10], ESPRIT [11], propagator
method [12], higher-order singular value decomposition (HOSVD)
[13–16] and optimization-based methods [17,18]. Unfortunately, the
algorithmic performances in [9,18] display degradation in the presence
of spatial colored noise. Several algorithms has been presented to deal
with the spatial colored noise. In term of the mechanism to suppress
the spatial colored noise, these algorithms can be grouped into two
classes. One is spatial cross-correlation methods [19–22], in which the
orthogonal characteristic of the colored noise between different
matched filters is explored. The other is temporal cross-correlation
methods [23], where the orthogonal characteristic of the colored noise
between different snapshots is exploited. In [19], a spatial cross-
correlation method is presented, but it is only feasible for three
transmit antennas configured MIMO radar. By dividing the transmit
array into two subarrays, an improved spatial cross-correlation scheme
is present in [20], which can eliminate the spatial colored noise for
three or more transmit antennas configuration. Another cross-correla-
tion scheme is derived in [21], whose transmit array and receive array
are both divided into two subarrays. It is e
ffective
to suppress the
spatial colored noise at the penalty of high-complexity computation, as
the two transmitter subarrays and receiver subarrays are well separated
and it required four dimensional angle estimation. To exploit the
structure inherent in the array data, a tensor-based spatial cross-
correlation method is proposed in [22]. Regretfully, a major weakness
in [23,22] is the virtual aperture loss, because the cross-correlation
http://dx.doi.org/10.1016/j.sigpro.2016.12.017
Received 30 August 2016; Received in revised form 10 December 2016; Accepted 16 December 2016
☆
Fully documented templates are available in the elsarticle package on CTAN http://www.ctan.org/tex-archive/macros/latex/contrib/elsarticle.
⁎
Corresponding author.
E-mail address: wfqltt@163.com (F. Wen).
Signal Processing 134 (2017) 261–267
Available online 21 December 2016
0165-1684/ © 2016 Elsevier B.V. All rights reserved.
MARK