Computationally effective spectral MUSIC algorithm
for monostatic MIMO radar using real
polynomial rooting
XU Liqin, LI Yong
School of Electronics and Information
Northwestern Polytechnical University
Xi’an, China
xuliqin@xupt.edu.cn, ruikel@nwpu.edu.cn
Abstract—A computationally effective real-valued variant of
multiple signal classification (MUSIC) algorithm for monostatic
multiple-input multiple-output (MIMO) radar is presented.
Reduced-dimension transformation is utilized to reduce the
dimension of the received data matrix at first, and then the
unitary transformation is employed to transform the complex
covariance matrix of the received data into a real-valued one. To
further reduce the computational complexity, a real polynomial
rooting technique is presented to determine the local maxima of
the MUSIC spectrum that corresponding to the DOAs of the
targets instead of the computationally-expensive spectrum search.
Simulations results demonstrate that the presented algorithm can
greatly reduce the computational complexity without sacrificing
the estimation accuracy.
Keywords—multiple input multiple output (MIMO) radar;
direction of arrival (DOA) estimation ; multiple signal classification
(MUSIC)
I. I
NTRODUCTION
Multiple-input multiple-output (MIMO) radar transmits
mutually orthogonal waveforms by using multiple transmit
antennas and receives the reflected signals by using multiple
receive antennas
[1]
. It possesses many potential advantages
over the traditional phase-array radar, such as higher accuracy
of angle estimation, more degree of freedom and better
parameter identification performance
[2]
. Direction of arrival
(DOA) estimation, as an important issue in MIMO radar, has
attracted considerable attention and intensive research
[3-7].
Multiple signal classification (MUSIC) is a widely used high-
resolution DOA estimation algorithm. However, it requires
time-consuming two-dimensional (2-D) spatial spectrum
search to obtain the estimation of DODs and DOAs of the
targets when applied to a MIMO radar system. High accuracy
of angle estimation requires a very fine grid search, resulting in
high computational costs. Also, the computational complexity
of MUSIC increases dramatically with the number of array
sensors, which is particularly prominent in the MIMO radar
system. Many scholars dedicated to reduce the computational
complexity of MUSIC for MIMO radar
[8-13]
. In [8], a reduced-
dimensional MUSIC algorithm is proposed. It only needs 1-D
spectrum search to obtain automatically paired DODs and
DOAs. In [9] and [10], the root-MUSIC algorithm is proposed
for bistatic MIMO radar, in which polynomial root finding
technique is utilized to obtain the DOAs and DODs of the
targets, completely avoiding the computational expensive
spectrum search. However, it still has relatively high
complexity as it needs to find the roots of a high-order
complex polynomial. Reduced complexity unitary root-
MUSIC algorithm is presented in [11], in which reduced-
dimension transformation and unitary transformation are
jointly used to reduce the computational complexity. In [12], a
beamspace root-MUSIC algorithm is presented for bistatic
MIMO radar. The computational burden is relieved by
mapping the received data into lower-dimensional beamspace.
In [13], a beamspace spectral MUSIC algorithm is proposed to
reduce the computational load by using a combination of
reduced-dimension transformation and beamspace processing.
However, it needs to know the priori information about the
rough location of the targets and a 1-D spectrum search is still
required.
In this paper, we present a computationally efficient
spectral MUSIC angle estimation algorithm for monostatic
MIMO radar. The algorithm firstly utilizes reduced-dimension
transformation to transform the observed data into a lower-
dimensional space and then employs unitary transformation to
reduce the computational load in the eigen-decomposition
stage. Finally, a real polynomial rooting technique is used
instead of the expensive spectrum search to determine the
local maxima of the MUSIC spectrum. The presented
algorithm can greatly reduce the computational load without
sacrificing the accuracy of angle estimation.
II. S
IGNAL
M
ODEL
Consider a narrowband monostatic MIMO radar system
composed of uniform linear arrays. Assume that the
transmit antennas and
N
receive antennas are spaced half-
wavelength apart. At the transmitting end, transmit antennas
transmit mutually orthogonal encoded pulse signals. Assume
that there are
non-coherent targets located in the
directions
12
,,,
P
θθ θ
!
, respectively. Echo signals reflected by
the
targets are received by the multiple antennas in the
receiver. After matched filtering, the observed data can be
expressed as
This work was supported by the National Natural Science foundation o
China (61602377)
2182
978-1-5386-3758-6/18/$31.00
c
2018 IEEE