Open AccessArticle

Noncircular Sources-Based Sparse Representation Algorithm for Direction of Arrival Estimation in MIMO Radar with Mutual Coupling

Weidong Zhou

Jing Liu

^*,

Pengxiang Zhu

Wenhe Gong

and

Jiaxin Hou

College of Automation, Harbin Engineering University, Harbin 150001, China

Author to whom correspondence should be addressed.

Algorithms 2016, 9(3), 61; https://doi.org/10.3390/a9030061

Submission received: 25 June 2016 / Revised: 1 September 2016 / Accepted: 2 September 2016 / Published: 8 September 2016

Download

Browse Figures

Versions Notes

Abstract

In this paper, a reweighted sparse representation algorithm based on noncircular sources is proposed, and the problem of the direction of arrival (DOA) estimation for multiple-input multiple-output (MIMO) radar with mutual coupling is addressed. Making full use of the special structure of banded symmetric Toeplitz mutual coupling matrices (MCM), the proposed algorithm firstly eliminates the effect of mutual coupling by linear transformation. Then, a reduced dimensional transformation is exploited to reduce the computational complexity of the proposed algorithm. Furthermore, by utilizing the noncircular feature of signals, the new extended received data matrix is formulated to enlarge the array aperture. Finally, based on the new received data, a reweighted matrix is constructed, and the proposed method further designs the joint reweighted sparse representation scheme to achieve the DOA estimation by solving the

l_{1}

-norm constraint minimization problem. The proposed method enlarges the array aperture due to the application of signal noncircularity, and in the presence of mutual coupling, the proposed algorithm provides higher resolution and better angle estimation performance than ESPRIT-like,

l_{1}

-SVD and

l_{1}

-SRDML (sparse representation deterministic maximum likelihood) algorithms. Numerical experiment results verify the effectiveness and advantages of the proposed method.

Keywords:

multiple-input multiple-output radar; direction of arrival estimation; mutual coupling; reweighted sparse representation; noncircular signal

1. Introduction

With orthogonal transmitted waveforms, multiple-input multiple-output (MIMO) radar has drawn increasing attention in the field of wireless communications. Compared with the conventional phased-array radar, MIMO radar owns a number of advantages, such as higher resolution and better parameter identifiability [1]. In addition, MIMO radar can be classified into the following two types: statistical MIMO radar and colocated MIMO radar. Colocated MIMO radar can achieve more degrees of freedom and higher spatial resolution, because its closely-spaced antennas form a virtual array with a large aperture. Colocated MIMO radar includes the monostatic one and the bistatic one. In bistatic MIMO radar, the transmit array and the receive array are separated away, while they are close to each other in monostatic MIMO radar. In this paper, the angle estimation problem is investigated in monostatic MIMO radar.

Parameter estimation is an important aspect in wireless communications and sensor array signal processing [2]. In order to estimate the direction of arrival (DOA), a large number of subspace-based algorithms have been proposed [3,4,5]. Recently, the emerging sparse representation (SR) and compressed sensing (CS) have attracted more and more attention in various fields [6,7,8], such as sensor array signal processing, imaging and image processing, and so forth. Furthermore, it has been verified that SR-based DOA estimation algorithms have remarkable advantages over other methods [9], for instance they adapt better to challenging circumstances and provide higher angle resolution. A few SR-based algorithms, such as

l_{1}

-SVD (singular value decomposition) [9] and

l_{1}

-SRACV (array covariance vectors) [10], have been proposed via sparse signal recovery. In [11], based on the array covariance vector, the revised real-valued (RV)

l_{1}

-SVD approach has better DOA estimation performance and lower computation burden than

l_{1}

-SVD [9] and RV

l_{1}

-SVD [12].

However, in practical situations, the sensor array manifold is often affected by the errors of unknown mutual coupling. When they are taken into account, the DOA estimation performance of the above-mentioned methods is seriously degraded. With mutual coupling, the subspace-based MUSIC-like (multiple signal classification-like) [13] algorithm is proposed. Besides, a ESPRIT-like (estimation of signal parameters via rotational invariance techniques-like) algorithm is presented in [14], which owns a low calculation burden. In [15], the SR-based revised

l_{1}

-SVD algorithm eliminates the errors of mutual coupling, and it achieves angle estimation by solving the problem of the

l_{1}

-norm constraint minimization. In [16], in the presence of mutual coupling, by introducing the refined deterministic maximum likelihood (DML) procedure, a covariance vector-based sparse representation algorithm is proposed with outstanding angle estimation performance. We call the method in [16]

l_{1}

-SRDML (sparse representation deterministic maximum likelihood).

In practical communication systems and radar applications, the complex noncircular sources are widely used, such as binary phase shift keying (BPSK), multiple amplitude shift keying (MASK) and unbalanced quadrature phase shift keying (UQPSK) modulated signals [17]. Noncircular signals can be used to enlarge the array aperture without extra antennas, which contributes to the improvement of angle estimation performance [18]. Exploiting the signal noncircularity, some subspace-based methods have been developed for angle estimation in MIMO radar [17,19,20]. They have verified that compared with traditional subspace-based algorithms, the methods that are based on noncircular signals can achieve better DOA estimation performance. However, these subspace-based approaches do not take the errors of unknown mutual coupling into account. Moreover, they perform poorly when experiencing challenging circumstances, such as fewer snapshots, which can be solved by sparse representation-based methods. For DOA estimation in MIMO radar with mutual coupling, to the best of our knowledge, up to now, there are no references about SR-based methods that exploit the signal noncircularity to extend the virtual array. Consequently, utilizing noncircular sources and the sparse representation framework, we aim to eliminate the effect of mutual coupling and then achieve a better DOA estimation.

In this paper, by using the noncircular signals, we propose a sparse representation-based DOA estimation algorithm in MIMO radar with mutual coupling. The new method firstly eliminates the effect of mutual coupling by extracting the information of mutual coupling coefficients into a diagonal matrix. Secondly, based on the reduced dimensional transformation and the characteristics of signal noncircularity, the new extended received data matrix is formulated to enlarge the array aperture. Finally, a reweighted matrix is constructed, based on which a joint reweighted

l_{1}

-norm minimization sparse representation framework is designed to obtain the DOAs. In the presence of mutual coupling, the proposed algorithm performs better than the conventional algorithms on account of the error elimination and the application of noncircular signals.

The rest of this paper is organized as follows. In Section 2, the MIMO radar system model with mutual coupling and noncircular signals is described. In Section 3, the implementation process of the proposed method is described in detail, including mutual coupling elimination, noncircular signal-based extended matrix construction and the joint reweighted sparse representation-based DOA estimation scheme. In Section 4, we give some related remarks and discussions regarding the parameter setting and the computational complexity of the proposed method. Then, in Section 5, simulations are implemented to verify the efficiency and the advantages of the proposed method. Finally, our conclusions are summarized in Section 6.

Notation:

{(\cdot)}^{H}

{(\cdot)}^{T}

{(\cdot)}^{*}

{(\cdot)}^{- 1}

E (\cdot)

and

\det (\cdot)

denote conjugate-transpose, transpose, conjugate, inverse, expectation and determinant operators, respectively. ⊗ and ⊙ denote the Kronecker product and the Khatri–Rao product, respectively.

I_{K}

denotes a

K \times K

dimensional unit matrix.

∥ \cdot ∥_{1}

∥ \cdot ∥_{2}

and

∥ \cdot ∥_{F}

denote the

l_{1}

-norm, the

l_{2}

-norm and the Frobenius-norm, respectively.

diag (\cdot)

denotes the diagonalization operation.

2. Problem Formulation

2.1. MIMO Radar System Model with Mutual Coupling

Consider a narrowband monostatic MIMO radar system, shown in Figure 1, whose transmit and receive arrays are both half-wavelength d spaced uniform linear arrays (ULAs). In the transmit array, M antennas transmit M orthogonal noncircular waveforms, such as BPSK modulated signals. In the receive array, N antennas are impinged by P uncorrelated targets that can be deemed as point scatterers in the far-field. Different from the mutually-separated arrays in bistatic MIMO radar, the transmitter and the receiver are closely located in monostatic radar, as shown in Figure 2. Since the transmit array and the receive array are collocated at the same site in monostatic MIMO radar systems [17], the echo signal travels the same path as the transmitted signal. Therefore, for the p-th target, the directions of departure of the transmitted signal and arrival of the echo from the target are the same [21], i.e.,

θ_{D O D} = θ_{D O A}

, both of which are denoted as DOA

θ_{p}, p = 1, 2, \dots P

. When mutual coupling is taken into account in both the transmit array and the receive array, at the receiver, the output of the matched filters at a certain snapshot is given by [16]:

x (t) = \tilde{A} s (t) + n (t)

(1)

where

x (t) \in C^{M N \times 1}

is the received data vector,

s (t) = {[s_{1} (t), s_{2} (t), \dots, s_{P} (t)]}^{T} \in C^{P \times 1}

is the complex-valued noncircular signal vector and

n (t) \in C^{M N \times 1}

is the additional Gaussian noise vector with zero-mean and covariance matrix

σ^{2} I_{M N}

, in which

σ^{2}

is the noise power. Meanwhile, the detailed expression of

\tilde{A}

is [14]:

\begin{matrix} \tilde{A} & = [C_{t} a_{t} (θ_{1}) \otimes C_{r} a_{r} (θ_{1}), \dots, C_{t} a_{t} (θ_{P}) \otimes C_{r} a_{r} (θ_{P})] \\ = (C_{t} \otimes C_{r}) [a_{t} (θ_{1}) \otimes a_{r} (θ_{1}), \dots, a_{t} (θ_{P}) \otimes a_{r} (θ_{P})] \end{matrix}

(2)

where

\tilde{A} \in C^{M N \times P}

a_{t} (θ_{p}) = {[1, e^{j π sin (θ_{p})}, e^{j π 2 sin (θ_{p})}, \dots, e^{j π (M - 1) sin (θ_{p})}]}^{T}

is the transmit steering vector and

a_{r} (θ_{p}) = {[1, e^{j π sin (θ_{p})}, e^{j π 2 sin (θ_{p})}, \dots, e^{j π (N - 1) sin (θ_{p})}]}^{T}

is the receive steering vector,

p = 1, 2, \dots P

C_{t}

and

C_{r}

are mutual coupling matrices of the transmit array and the receive array, respectively.

C_{t}

and

C_{r}

can be modeled as banded symmetric Toeplitz matrices, which are expressed as [22,23,24]:

\begin{matrix} C_{i} = [\begin{matrix} c_{i 0} & c_{i 1} & \dots & c_{i k} & \dots & 0 \\ c_{i 1} & c_{i 0} & c_{i 1} & \dots & ⋱ & ⋮ \\ ⋮ & c_{i 1} & c_{i 0} & ⋱ & ⋱ & c_{i k} \\ c_{i k} & \dots & ⋱ & ⋱ & c_{i 1} & ⋮ \\ ⋮ & ⋱ & \dots & c_{i 1} & c_{i 0} & c_{i 1} \\ 0 & \dots & c_{i k} & \dots & c_{i 1} & c_{i 0} \end{matrix}] \end{matrix}

(3)

where

i = t, r

c_{i j}

is the non-zero mutual coupling coefficient,

j = 0, 1, . . ., k

. In Figure 1, the coupling effects of the m-th and the n-th elements in the transmit and the receive arrays are described, respectively.

β_{(m, q)}

is the coupling contribution from the q-th to the m-th elements in transmit array, and

{\bar{β}}_{(n, q^{'})}

is the contribution from the

q^{'}

-th to the n-th elements in the receive array,

1 \leq m \leq M

1 \leq n \leq N

, and

| q - m | \leq k

| q^{'} - n | \leq k

β_{(m, q)}

and

{\bar{β}}_{(n, q^{'})}

contain the information of

c_{t | q - m |}

and

c_{r | q^{'} - n |}

, respectively. Mutual coupling coefficients are factors related to the distance between the two antennas [23], and for the

k + 1

non-zero coefficients, they satisfy

0 < | c_{i k} | < . . . < | c_{i 1} | < | c_{i 0} | = 1

. Besides, for the ULA coupling model, just a few mutual coupling coefficients are non-zero ones [22,23,24], and without loss of generality,

2 k < \min {M, N}

is assumed. Then, by collecting J snapshots, the received data in MIMO radar with mutual coupling is represented as follows:

X = C A S + N

(4)

where

X \in C^{M N \times J}

is the data matrix,

C = (C_{t} \otimes C_{r}) \in C^{M N \times M N}

is the transmit-receive mutual coupling matrix.

A = [a (θ_{1}), \dots, a (θ_{P})] \in C^{M N \times P}

is the steering matrix, with the steering vector

a (θ_{p}) = a_{t} (θ_{p}) \otimes a_{r} (θ_{p}), p = 1, 2, \dots P

S = [s (t_{1}), \dots, s (t_{J})] \in C^{P \times J}

and

N = [n (t_{1}), \dots, n (t_{J})] \in C^{M N \times J}

are the noncircular signal matrix and the complex Gaussian white noise matrix, respectively.

2.2. Noncircular Signals

For complex random variables and vectors, circularity is an important property [18]. Making full use of the properties of noncircular signals, the array aperture can be enlarged. The invariance specialityof rotation can be used to judge if a complex random sequence is noncircular [18]. More specifically, if stationary complex signal sequence

\bar{q}

is circular, it accords with

E [\bar{q} {\bar{q}}^{H}] \neq 0

and

E [\bar{q} {\bar{q}}^{T}] = 0

. By comparison, if the sequence is noncircular, then

E [\bar{q} {\bar{q}}^{H}] \neq 0

and

E [\bar{q} {\bar{q}}^{T}] \neq 0

, which can be utilized to increase the number of effective array elements. In addition, for zero-mean stationary complex signal sequence

\bar{q}

, it holds the property that:

E [\bar{q} {\bar{q}}^{T}] = ρ e^{j ψ} E [\bar{q} {\bar{q}}^{H}]

(5)

where

0 \leq ρ \leq 1

and ψ are noncircular rate and noncircular phase angle, respectively, and ρ depends on the signal modulation type. In this paper, we consider the signals with maximum noncircular rate, namely,

ρ = 1

s (t)

, such as BPSK and MASK modulated signals that are widely used in MIMO radar systems. It has been pointed out that the noncircular signal vector

s (t)

in Equation (1) can be rewritten as [17,25]:

s (t) = Λ_{φ} s_{r} (t)

(6)

where

s_{r} (t) = {[s_{r 1} (t), s_{r 2} (t), \dots, s_{r P} (t)]}^{T}

is the real-valued part of noncircular signal

s (t)

, and

Λ_{φ} = diag ([e^{j φ_{1}}, e^{j φ_{2}}, \dots, e^{j φ_{P}}])

is the noncircular phase matrix. Based on the composition and structure of

s (t)

, for the p-th target with

t = 1, 2, \dots, J

E [s_{p} s_{p}^{T}] = e^{j 2 φ_{p}} E [s_{r p} s_{r p}^{T}] = e^{j 2 φ_{p}} E [s_{p} s_{p}^{H}]

can be obtained. In addition, each element in

s (t)

accords with Equation (5), i.e.,

E [s_{p} s_{p}^{T}] = e^{j ψ} E [s_{p} s_{p}^{H}]

with

ρ = 1

[26]. This indicates that the relationship between

φ_{p}

and

ψ_{p}

can be written as

φ_{p} = ψ_{p} / 2

. With the application of signal noncircularity,

X

in Equation (4) is expressed as:

X = C A Λ_{φ} S_{r} + N

(7)

where

S_{r} = [s_{r} (t_{1}), \dots, s_{r} (t_{J})]

is a real-valued matrix.

3. The Proposed Algorithm

3.1. Mutual Coupling Elimination

To successfully implement the noncircular source-based sparse representation method, we firstly eliminate the effect of mutual coupling. Exploiting the structure of mutual coupling matrices

C_{t}

and

C_{r}

, we define two selection matrices as follows:

\begin{matrix} Γ_{1} & = [0_{(M - 2 k) \times k} I_{(M - 2 k) \times (M - 2 k)} 0_{(M - 2 k) \times k}] \\ Γ_{2} & = [0_{(N - 2 k) \times k} I_{(N - 2 k) \times (N - 2 k)} 0_{(N - 2 k) \times k}] \end{matrix}

(8)

where

Γ_{1} \in C^{(M - 2 k) \times M}

, and

Γ_{2} \in C^{(N - 2 k) \times N}

Γ_{1}

and

Γ_{2}

can be used to choose the central rows of the mutual coupling matrices and then transform the transmit-receive mutual coupling matrix

C

into a diagonal one. Thus, let

z_{p} = e^{j π sin (θ_{p})}

, by multiplying the selection matrix

Γ_{1}

on the left side of the transmit steering vector with mutual coupling; the transformation of the i-th element in

Γ_{1} C_{t} a_{t} (θ_{p}) \in C^{(M - 2 k) \times 1}

can be derived as:

\begin{matrix} [Γ_{1} C_{t} a_{t} (θ_{p})] (i) & = c_{t k} z_{p}^{i - 1} + c_{t (k - 1)} z_{p}^{i} + \dots + z_{p}^{i + k - 1} + \dots + c_{t k} z_{p}^{i + k - 1 + k} \\ = (\sum_{j = - k}^{k} c_{t | j |} z_{p}^{j + k}) z_{p}^{i - 1} \\ = ω_{t p} z_{p}^{i - 1} \end{matrix}

(9)

where

ω_{t p}

is a scalar that just contains the information of

θ_{p}

and mutual coupling in the transmit array. Let

\bar{M} = M - 2 k

; thus, for

i = 1, 2, \dots, \bar{M}

, we have

Γ_{1} C_{t} a_{t} (θ_{p}) = ω_{t p} {\tilde{a}}_{t}

, in which

{\tilde{a}}_{t}

is the new transmit steering vector and can be expressed as

{\tilde{a}}_{t} = [1, z_{p}, z_{p}^{2} \dots, z_{p}^{\bar{M} - 1}]

p = 1, 2, \dots, P

. Similar to Equation (9), we use the other selection matrix

Γ_{2}

to extract the mutual coupling coefficients in

C_{r} a_{r} (θ_{p})

into a scalar, that is:

\begin{matrix} [Γ_{1} C_{r} a_{r} (θ_{p})] (i) & = c_{r k} z_{p}^{i - 1} + c_{r (k - 1)} z_{p}^{i} + \dots + z_{p}^{i + k - 1} + \dots + c_{r k} z_{p}^{i + k - 1 + k} \\ = (\sum_{j = - k}^{k} c_{r | j |} z_{p}^{j + k}) z_{p}^{i - 1} \\ = ω_{r p} z_{p}^{i - 1} \end{matrix}

(10)

where

p = 1, 2, \dots, P

. The value of

ω_{r p}

depends on the mutual coupling coefficients and

θ_{p}

. Then, for

i = 1, 2, \dots, \bar{N}

, we further have

Γ_{2} C_{r} a_{r} (θ_{p}) = ω_{r p} {\tilde{a}}_{r}

with

{\tilde{a}}_{r} = [1, z_{p}, z_{p}^{2} \dots, z_{p}^{\bar{N} - 1}]

being the new receive steering vector. As a result, exploiting the property of the Kronecker product operator, let

Γ = Γ_{1} \otimes Γ_{2} \in C^{\bar{M} \bar{N} \times M N}

be the selection matrix of transmit-receive array; we deduce that:

\begin{matrix} Γ C a (θ_{p}) & = [Γ_{1} C_{t} a_{t} (θ_{p})] \otimes [Γ_{2} C_{r} a_{r} (θ_{p})] \\ = ω_{t p} ω_{r p} \tilde{a} (θ_{p}) \end{matrix}

(11)

where

\tilde{a} (θ_{p}) = {\tilde{a}}_{t} (θ_{p}) \otimes {\tilde{a}}_{r} (θ_{p}) \in C^{\bar{M} \bar{N} \times 1}

is the new transmit-receive steering vector. Thus, for J snapshots, the data matrix

X

in Equation (7) transfers into:

\bar{X} = \bar{A} D Λ_{φ} S_{r} + \bar{N}

(12)

where

\bar{A} = [\tilde{a} (θ_{1}), \tilde{a} (θ_{2}), \dots, \tilde{a} (θ_{P})] \in C^{\bar{M} \bar{N} \times P}

and

\bar{N} = Γ N \in C^{\bar{M} \bar{N} \times J}

are the new transmit-receive steering matrix and noise matrix, respectively; in addition,

D = diag [ω_{t 1} ω_{r 1}, ω_{t 2} ω_{r 2}, \dots, ω_{t P} ω_{r P}] \in C^{P \times P}

. It can be observed in Equation (12) that non-zero mutual coupling coefficients are extracted into a diagonal matrix

D

. Hence, for the new steering matrix

\bar{A}

, the errors of mutual coupling are eliminated.

3.2. Noncircular Signal-Based Extended Matrix Construction

Based on the new received data matrix

\bar{X}

, a reduced dimensional transformation is introduced in the following. For the p-th target, the transmit-receive steering vector

\tilde{a} (θ_{p})

can be expressed as:

\begin{matrix} \tilde{a} (θ_{p}) & = {(1, \dots, z_{p}^{\bar{N} - 1}; z_{p}, \dots, z_{p}^{\bar{N}}; \dots; z_{p}^{\bar{M} - 1}, \dots, z_{p}^{\bar{M} + \bar{N} - 2})}^{T} \end{matrix}

(13)

where

p = 1, 2, \dots, P

. Many repeated terms are contained in

\tilde{a} (θ_{p})

, and the non-repeated elements can make up another vector

b (θ_{p}) = {[1, \exp (j π \sin θ_{p}), \dots, \exp (j π (\bar{M} + \bar{N} - 2) \sin θ_{p})]}^{T} \in C^{(\bar{M} + \bar{N} - 1) \times 1}

. The relationship between

\tilde{a} (θ_{p})

and

b (θ_{p})

can be derived as

\tilde{a} (θ_{p}) = Gb (θ_{p})

, in which

G

is represented as:

\begin{matrix} G = {[L_{0}^{T}, L_{1}^{T}, \dots, L_{\bar{M} - 1}^{T}]}^{T} \end{matrix}

(14)

where

G \in C^{\bar{M} \bar{N} \times (\bar{M} + \bar{N} - 1)}

, and

L_{m} = [0_{\bar{N} \times m}, I_{\bar{N}}, 0_{\bar{N} \times (\bar{M} - m - 1)}] \in C^{\bar{N} \times (\bar{M} + \bar{N} - 1)}

for

m = 0, 1, . . ., \bar{M} - 1

. Thus, based on

G

, a reduced dimensional transformation matrix can be constructed as

R = {(G^{H} G)}^{(- \frac{1}{2})} G^{H}

, then

\bar{X}

in Equation (12) turns into:

\begin{matrix} Y & = {(G^{H} G)}^{(- \frac{1}{2})} G^{H} G B D Λ_{φ} S_{r} + R \bar{N} \\ = F B D Λ_{φ} S_{r} + {\bar{N}}_{d} \end{matrix}

(15)

where

Y

and

{\bar{N}}_{d} = R \bar{N}

are the new reduced dimensional data matrix and the new noise matrix, respectively.

F = {(G^{H} G)}^{(\frac{1}{2})} \in C^{(\bar{M} + \bar{N} - 1) \times (\bar{M} + \bar{N} - 1)}

can be directly calculated as:

\begin{matrix} F = diag [1, \sqrt{2}, . . ., \underset{| \bar{M} - \bar{N} | + 1}{\underset{︸}{\min (\sqrt{\bar{M}}, \sqrt{\bar{N}}), . . ., \min (\sqrt{\bar{M}}, \sqrt{\bar{N}})}}, . . . ., \sqrt{2}, 1] \end{matrix}

(16)

By the transformation, the dimension of the data matrix is reduced from

\bar{M} \bar{N} \times J

(\bar{M} + \bar{N} - 1) \times J

. In addition, as

R R^{H} = I_{\bar{M} + \bar{N} - 1}

, the transformation does not bring in the additional spatial colored noises. In order to enlarge the array aperture by exploiting the noncircular signals, the new received data matrix

Y \in C^{(\bar{M} + \bar{N} - 1) \times J}

can be extended as

Z = {[Y^{T}, Y^{H}]}^{T} \in C^{2 (\bar{M} + \bar{N} - 1) \times J}

. Since

F

D

and

S_{r}

are real-valued matrices,

Z

is expressed as:

\begin{matrix} Z = [\begin{matrix} F B D Λ_{φ} \\ F B^{*} D Λ_{φ}^{*} \end{matrix}] S_{r} + [\begin{matrix} {\bar{N}}_{d} \\ {\bar{N}}_{d}^{*} \end{matrix}] \end{matrix}

(17)

3.3. Joint Reweighted Sparse Representation-Based DOA Estimation Scheme

In order to successfully apply the sparse representation theory to estimate the DOAs, we bring in the SVD technique of

Z

, i.e.,

Z = U Λ V^{H}

, where

Λ = diag (ω_{1}, ω_{2}, \dots, ω_{\bar{M} + \bar{N} - 1})

, and

ω_{1} \geq ω_{2} \geq \dots \geq ω_{\bar{M} + \bar{N} - 1}

are singular values. Let

V_{s}

be composed of the P vectors in

V

that correspond to

(ω_{1}, ω_{2}, \dots, ω_{P})

. With

V_{s}

, the dimension of the observation matrix can be further reduced [9], which is conducive to the signal reconstruction, that is:

\begin{matrix} Z_{v} = [\begin{matrix} F B D Λ_{φ} S_{r} V_{s} \\ F B^{*} D Λ_{φ}^{*} S_{r} V_{s} \end{matrix}] + [\begin{matrix} {\bar{N}}_{d} V_{s} \\ {\bar{N}}_{d}^{*} V_{s} \end{matrix}] = [\begin{matrix} \bar{B} \bar{T} \\ \tilde{B} \tilde{T} \end{matrix}] + [\begin{matrix} {\bar{N}}_{d s} \\ {\tilde{N}}_{d s} \end{matrix}] \end{matrix}

(18)

where

Z_{v} = Z V_{s} \in C^{2 (\bar{M} + \bar{N} - 1) \times P}

\bar{B} = F B

\bar{T} = D Λ_{φ} S_{r} V_{s}

\tilde{B} = F B^{*}

\tilde{T} = D Λ_{φ}^{*} S_{r} V_{s}

{\bar{N}}_{d s}

and

{\tilde{N}}_{d s}

are the new noise matrices.

Z_{v}

can be divided into the following two parts; the first

(\bar{M} + \bar{N} - 1)

rows in

Z_{v}

make up

Z_{v 1} \in C^{(\bar{M} + \bar{N} - 1) \times P}

, and the remaining rows compose

Z_{v 2} \in C^{(\bar{M} + \bar{N} - 1) \times P}

, namely

Z_{v} = {[Z_{v 1}^{T}, Z_{v 2}^{T}]}^{T}

. By constructing two complete dictionaries

{\bar{B}}_{\hat{θ}} = F [b ({\hat{θ}}_{1}), b ({\hat{θ}}_{2}), \dots, b ({\hat{θ}}_{L})]

and

{\tilde{B}}_{\hat{θ}} = F [b^{*} ({\hat{θ}}_{1}), b^{*} ({\hat{θ}}_{2}), \dots, b^{*} ({\hat{θ}}_{L})]

with

{{\hat{θ}}_{l}}_{l = 1}^{L} (L ≫ P)

being the discretized sampling grids of all potential DOAs, the corresponding sparse representation models can be separately formulated as:

\begin{matrix} Z_{v 1} & = {\bar{B}}_{\hat{θ}} {\bar{T}}_{\hat{θ}} + {\bar{N}}_{d s} \\ Z_{v 2} & = {\tilde{B}}_{\hat{θ}} {\tilde{T}}_{\hat{θ}} + {\tilde{N}}_{d s} \end{matrix}

(19)

where

{\bar{T}}_{\hat{θ}} \in C^{L \times P}

and

{\tilde{T}}_{\hat{θ}} \in C^{L \times P}

hold the same row support with

\bar{T}

and

\tilde{T}

, respectively. For the two models in Equation (19), if they are solved independently by conventional sparse representation methods that are based on the

l_{1}

-norm penalty [9], the signal noncircularity cannot be utilized to extend the array aperture. In order to make the best of signal noncircularity, we design a joint sparse vector

r \in C^{L \times 1}

whose k-th element is

r (k) = \sqrt{{({\bar{T}}_{\hat{θ}}^{(l_{2})} (k))}^{2} + {({\tilde{T}}_{\hat{θ}}^{(l_{2})} (k))}^{2}}

, then the sparsities of

{\bar{T}}_{\hat{θ}}

and

{\tilde{T}}_{\hat{θ}}

are combined. In

r (k)

{\bar{T}}_{\hat{θ}}^{(l_{2})} (k)

and

{\tilde{T}}_{\hat{θ}}^{(l_{2})} (k)

are defined as

{\bar{T}}_{\hat{θ}}^{(l_{2})} (k) = ∥ {\bar{T}}_{\hat{θ}} (k, :) ∥_{2}

and

{\tilde{T}}_{\hat{θ}}^{(l_{2})} (k) = ∥ {\tilde{T}}_{\hat{θ}} (k, :) ∥_{2}

, respectively. More specifically,

{\bar{T}}_{\hat{θ}}^{(l_{2})} (k)

is equal to the

l_{2}

-norm of the k-th row in

{\bar{T}}_{\hat{θ}}

; similarly,

{\tilde{T}}_{\hat{θ}}^{(l_{2})} (k)

is equal to the

l_{2}

-norm of the k-th row in

{\tilde{T}}_{\hat{θ}}

. Let

\hat{r}

represent the solution vector of the sparse signal reconstruction,

\hat{r} (k) \geq r (k)

k = 1, 2, \dots, L

. As a result, based on the

l_{1}

-norm constrained minimization, a joint sparse representation scheme is designed as follows:

\min ∥ \hat{r} ∥_{1}, s . t . ∥ Z_{v 1} - {\bar{B}}_{\hat{θ}} {\bar{T}}_{\hat{θ}} ∥_{F}^{2} \leq η_{1} ∥ Z_{v 2} - {\tilde{B}}_{\hat{θ}} {\tilde{T}}_{\hat{θ}} ∥_{F}^{2} \leq η_{2}

(20)

where

η_{1}

and

η_{2}

are regularization parameters that set the error amount. To make the

l_{1}

-norm penalty better approximate the

l_{0}

-norm penalty, we introduce the reweighted matrix in the following. Let

U_{n}

be the noise subspace and composed of the

(\bar{M} + \bar{N} - 1 - P)

vectors in

U_{n}

that correspond to

(ω_{P + 1}, ω_{P + 2}, \dots, ω_{\bar{M} + \bar{N} - 1})

. According to the subspace principle [13], for the true target

θ_{p}

{[{(ω_{t p} ω_{r p} e^{j φ_{p}} F b (θ_{p}))}^{T}, {(ω_{t p} ω_{r p} e^{- j φ_{p}} F b {(θ_{p})}^{*})}^{T}]}^{T}

Z

is orthogonal to the noise subspace

U_{n}

, that is:

\begin{matrix} Q (θ_{p}, φ_{p}, ω_{t p}, ω_{r p}) & = [ω_{t p} ω_{r p} e^{j φ_{p}} ω_{t p} ω_{r p} e^{- j φ_{p}}] \bar{Q} (θ_{p}) [\begin{matrix} ω_{t p} ω_{r p} e^{j φ_{p}} \\ ω_{t p} ω_{r p} e^{- j φ_{p}} \end{matrix}] \to 0 \end{matrix}

(21)

where:

\begin{matrix} \bar{Q} (θ_{p}) = {[\begin{matrix} F b (θ_{p}) 0_{(\bar{M} + \bar{N} - 1) \times 1} \\ 0_{(\bar{M} + \bar{N} - 1) \times 1} F b^{*} (θ_{p}) \end{matrix}]}^{H} U_{n} U_{n}^{H} [\begin{matrix} F b (θ_{p}) 0_{(\bar{M} + \bar{N} - 1) \times 1} \\ 0_{(\bar{M} + \bar{N} - 1) \times 1} F b^{*} (θ_{p}) \end{matrix}] \end{matrix}

(22)

Since

Q (θ_{p}, φ_{p}, ω_{t p}, ω_{r p}) \to 0

ω_{t p}

ω_{r p}

and

e^{j φ_{p}}

are all non-zero scalars and the rank of

U_{n} U_{n}^{H}

satisfies

rank (U_{n} U_{n}^{H}) > 2

, for the true DOA

θ_{p}

, the rank reduction of

\bar{Q} (θ_{p})

will take place, that is

\det [\bar{Q} (θ_{p})] \to 0

. Consequently, a reweighted vector is designed as:

\begin{matrix} w (i) = \det [\bar{Q} ({\hat{θ}}_{i})] \end{matrix}

(23)

for

i = 1, 2, \dots, L

. Thus, the reweighted matrix is:

\begin{matrix} W_{r} = diag (w) / \max (w) \end{matrix}

(24)

where

W_{r} \in C^{L \times L}

is a diagonal matrix. Large weights in

W_{r}

can punish the entries that are more likely to be zeros in

w

, while small weights reserve the larger entries. Hence,

W_{r}

can enhance the solution of the sparse recovery problem in Equation (20). Then, for DOA estimation in the presence of mutual coupling, the noncircular source-based joint reweighted sparse representation framework transforms into:

\min ∥ W_{r} \hat{r} ∥_{1}, s . t . ∥ Z_{v 1} - {\bar{B}}_{\hat{θ}} {\bar{T}}_{\hat{θ}} ∥_{F} \leq \sqrt{η_{1}} ∥ Z_{v 2} - {\tilde{B}}_{\hat{θ}} {\tilde{T}}_{\hat{θ}} ∥_{F} \leq \sqrt{η_{2}}

(25)

Effective SOC (second order cone) programming software packages, such as SeDuMi [27] and CVX [28], can be used to solve Equation (25). As a result, by plotting

\hat{r}

, the DOA estimation is achieved.

4. Related Remarks

Remark 1.

In the problem of the sparse signal reconstruction that is based on the

l_{1}

-norm minimization, the chosen values of the regularization parameters

η_{1}

and

η_{2}

are important for the estimation accuracy. In the proposed algorithm, noise matrix

N

in Equation (4) is complex white Gaussian with zero mean. By mutual coupling elimination and reduced dimensional transformation, the corresponding selection matrix Γ and transformation matrix

R

transform

N

into

{\bar{N}}_{d}

. According to the designed matrix structure of Γ and

R

Γ Γ^{H} = I_{\bar{M} \bar{N}}

and

R R^{H} = I_{\bar{M} + \bar{N} - 1}

can be obtained.

Based on the invariance speciality of linear transformation in an asymptotic normal distribution [29], if n is asymptotic normal with zero-mean and covariance matrix

σ^{2} I_{M N}

R Γ n

complies with a zero-mean asymptotic normal distribution with covariance matrix

R Γ (σ^{2} I_{M N}) {(R Γ)}^{H} = σ^{2} I_{\bar{M} + \bar{N} - 1}

. Therefore,

{\bar{N}}_{d} = [{\bar{n}}_{d 1}, {\bar{n}}_{d 2}, \dots, {\bar{n}}_{d J}]

in Equation (15) is a zero-mean complex Gaussian white noise matrix.

In [9], it has been verified that if the noise

{\bar{n}}_{d}

is independent and identically distributed (i.i.d.) Gaussian, then

∥ {\bar{N}}_{d s} ∥_{F}^{2}

has approximately a chi-square distribution with

(\bar{M} + \bar{N} - 1) P

degrees of freedom upon normalization by the variance of

{\bar{n}}_{d}

, where

{\bar{N}}_{d s} = {\bar{N}}_{d} V_{s}

with

V_{s}

being the signal subspace of

Z

[9]. As a result,

η_{1}

and

η_{2}

can be chosen as the upper limit value of

∥ {\bar{N}}_{d s} ∥_{F}^{2}

and

∥ {\tilde{N}}_{d s} ∥_{F}^{2}

upon

σ^{2}

, with a high probability

1 - ε

confidence interval, and

ε = 0.001

is enough. In addition,

σ^{2}

can be estimated by the average of the squares of

2 (\bar{M} + \bar{N} - 1) P

smallest singular values of

Z

. By using MATLAB software, the function

c h i 2 i n v (1 - ε, (\bar{M} + \bar{N} - 1) P) σ^{2}

can be used to compute the regularization parameters.

Remark 2.

The computational complexity of the proposed algorithm mainly focuses on eliminating the effect of mutual coupling, constructing the reweighted matrix and obtaining the sparse solution from Equation (25), which require

O {\bar{M} \bar{N} M N J}

O {[4 (\bar{M} + \bar{N} - 1) (2 (\bar{M} + \bar{N} - 1) - P) + 4 (\bar{M} + \bar{N} - 1) (2 (\bar{M} + \bar{N} - 1) - P) + 8 (\bar{M} + \bar{N} - 1)] \bar{L}}

and

O {{\bar{L}}^{3} P}

calculation burden, respectively, where

\bar{L}

is the number of the discretized sampling grids. Therefore, the total computational complexity of the proposed algorithm is

O {\bar{M} \bar{N} M N J + [4 (\bar{M} + \bar{N} - 1) (2 (\bar{M} + \bar{N} - 1) - P + 1)] \bar{L} + {\bar{L}}^{3} P}

. Although the constructions of the extended data matrix and the reweighted matrix increase the computational complexity, they enlarge the virtual array aperture and enhance the solution of the sparse reconstruction. As a result, the angle estimation performance of the proposed method is improved.

Remark 3.

The prior knowledge of the number of emitting sources is important. This is because in the absence of a perfect knowledge of the target number P, either adding spurious sources or missing actual signals may lead to the deviations of the signal and the noise subspaces, which further cause the deviations of extended observation matrix

Z_{v}

in Equation (18) and reweighted matrix

W_{r}

in Equation (24). Therefore, the selected value of P affects the performance of the proposed algorithm. Fortunately, when P is unknown in practical applications, some effective methods can be used to estimate it, such as the Akaike information criterion (AIC) and the minimum description length (MDL) [30]. In the proposed algorithm, the number of emitting sources P is assumed to be known.

Remark 4.

After performing the mutual coupling elimination and the reduced dimensional transformation, the received data matrix

X

in Equation (4) turns into

Y

in Equation (15), and the number of effective antenna elements reduces from

(M + N - 1)

(\bar{M} + \bar{N} - 1)

. Then, by using the real-valued signals that satisfy

S_{r} = S_{r}^{*}

extracted from the noncircular signals, the extended matrix

Z

with the dimension

2 (\bar{M} + \bar{N} - 1) \times J

is constructed in Equation (17). It implies that the extended data matrix

Z

corresponds to a virtual array, whose virtual antenna number is twice that of the array corresponding to

Y

. As the exploitation of the noncircularity of emitted signals enlarges the effective array aperture, the more diversity gain can be obtained in the proposed algorithm. In addition, the reweighted matrix enhances the sparse solution. Thus, the proposed algorithm considerably improves DOA estimation performance in the presence of mutual coupling.

Remark 5.

Mutual coupling coefficients are related to the distance between two antennas. For uniform linear arrays (ULAs) of the considered MIMO radar system in the proposed algorithm, the mutual coupling coefficients between two antenna elements satisfy

c_{t (q, m)} = c_{t (m, q)} = c_{t | q - m |}

for the transmit array and

c_{r (q^{'}, n)} = c_{r (n, q^{'})} = c_{r | q^{'} - n |}

for the receive array [22], in which

1 \leq q, m \leq M

and

1 \leq q^{'}, n \leq N

. Consequently, all of the effects of mutual coupling in ULA can be modeled as a banded symmetric Toeplitz matrix in Equation (3) [23], based on which selection matrices are constructed in Equation (8) to eliminate the mutual coupling. Mutual coupling matrix in Equation (3) is modeled under the circumstance of symmetric coupling in ULA. Hence, the proposed algorithm is not suitable for asymmetric coupling.

5. Simulation Results

In this section, using the ESPRIT-like [14] algorithm, the

l_{1}

-SVD [15] algorithm and the

l_{1}

-SRDML [16] algorithm for comparison, some simulation experiments are implemented to demonstrate the efficiency and the advantages of the proposed algorithm. The signal-to-noise ratio (SNR) is defined as SNR =

10 \log_{10} {(| | C A S | |}_{F}^{2} / {| | N | |}_{F}^{2}

). Let

{\hat{θ}}_{p}^{i}

be the estimation of the true DOA

θ_{p}

for the i-th Monte Carlo trial; the root mean square error (RMSE) of angle estimation is defined as:

(1 / P) \sum_{p = 1}^{P} \sqrt{(1 / Q) \sum_{i = 1}^{Q} {({\hat{θ}}_{p}^{i} - θ_{p})}^{2}}

(26)

where Q is the total number of the Monte Carlo trials, which is

Q = 200

in the simulations. Consider a narrowband monostatic MIMO radar system, whose arrays are both half-wavelength spaced ULAs with the effects of mutual coupling, and the numbers of transmit and receive antennas are M and N, respectively. Let

K = k + 1

be the nonzero mutual coupling coefficient number; two different mutual coupling cases of the transmitter and the receiver are considered in the following simulations: (1)

K = 2

with the nonzero mutual coupling coefficients being

[c_{t 0}, c_{t 1}] = [1, 0.1844 + j 0.0427]

and

[c_{r 0}, c_{r 1}] = [1, 0.0522 - j 0.1049]

; (2)

K = 3

with

[c_{t 0}, c_{t 1}, c_{t 2}] = [1, 0.6 + j 0.2, 0.02 + j 0.13]

and

[c_{r 0}, c_{r 1}, c_{r 2}] = [1, 0.5 + j 0.1, 0.01 + j 0.05]

. The number of uncorrelated targets P is assumed to be known. Besides, the confidence interval is set to

0.99

, and the discretized grids are uniform with

{0.1}^{\circ}

sampling from

- 90^{\circ}

90^{\circ}

, for the proposed method, as well as

l_{1}

-SVD and

l_{1}

-SRDML algorithms.

When performing the experiments, we solve the signal reconstruction problem of the noncircular source-based joint reweighted sparse representation framework in Equation (25). The sparse solution vector

\hat{r}

is obtained by using the SOC programming software packages. When there are true DOAs of targets located at certain points of the complete dictionary, the corresponding elements in

\hat{r}

are non-zero, and the rest are zero. Therefore, the true DOA locations in dictionary correspond to the largest P values of

| \hat{r} |

. In the simulations, the functional form

10 \log_{10} [| \hat{r} | / \max (| \hat{r} |)]

is used to plot the spatial spectrum of angle estimation. Then, by finding P peaks in the spectrum, DOA estimation is achieved.

Figure 3 depicts the spatial spectrum of the proposed method for the number of targets being

P = 3

and

P = 5

, respectively, where

M = N = 6

J = 200

, SNR is fixed at SNR = 0 dB and mutual coupling Case (1) is considered. For

P = 3

, the DOAs of the uncorrelated targets are

θ_{1} = - 10^{\circ}

θ_{2} = 0^{\circ}

θ_{3} = 10^{\circ}

. For

P = 5

, they are

θ_{1} = - 20^{\circ}

θ_{1} = - 10^{\circ}

θ_{1} = 0^{\circ}

θ_{2} = 10^{\circ}

and

θ_{3} = 20^{\circ}

, respectively. From Figure 3, it can be observed that the spatial spectrum peaks of the proposed method are sharp, and the sidelobe suppression is low. In addition, for different target numbers P, the spectrum maintains the accuracy of DOA estimation. This indicates that the proposed method is effective to perform DOA estimation accurately.

Figure 4 shows the RMSE of DOA estimation versus SNR in different methods with mutual coupling Case (1), where

M = N = 6

J = 200

and three targets with

θ_{1} = - 10^{\circ}

θ_{2} = 0^{\circ}

θ_{3} = 10^{\circ}

are considered. Additionally, the values of RMSE are computed by Equation (26) with Q trials. As can be seen in Figure 4,

l_{1}

-SRDML owns a lower RMSE than the ESPRIT-like and

l_{1}

-SVD algorithms, because the deterministic maximum likelihood procedure improves the estimation accuracy. In addition, the proposed method provides the best DOA estimation performance in all SNR regions, for the reason that the noncircularity-based joint reweighted sparse representation scheme enlarges the virtual array aperture and enhances the sparse solution.

Figure 5 illustrates the RMSE of DOA estimation versus SNR in different methods with mutual coupling Case (2), where

M = N = 7

J = 200

, and there are three uncorrelated targets located at

θ_{1} = - 15^{\circ}

θ_{2} = 0^{\circ}

θ_{3} = 15^{\circ}

. From Figure 5, it is clear that the performance of

l_{1}

-SVD is inferior to the other algorithms, and the proposed method has the lowest RMSE in all analyzed methods. This means that when the number of nonzero mutual coupling coefficients increases, the proposed method still has superior DOA estimation performance.

Figure 6 shows the RMSE of DOA estimation versus snapshots of different methods with mutual coupling Case (1), where

M = N = 6

, SNR = 0 dB, there are three uncorrelated targets with DOAs being

θ_{1} = - {11.5}^{\circ}

θ_{2} = 0^{\circ}

θ_{3} = {11.5}^{\circ}

and the number of snapshots varies from

J = 50

J = 550

. It can be seen from Figure 6 that the proposed method performs well in the case of fewer snapshots, and with obvious superiority over the ESPRIT-like,

l_{1}

-SVD and

l_{1}

-SRDML algorithms, the proposed method performs the best in all pf the ranges of snapshots on account of the application of the reweighted matrix and the noncircularity of emitted signals.

Figure 7 demonstrates the target resolution probability of different methods versus SNR with mutual coupling Case (1), where

M = N = 6

J = 100

and the DOAs of three targets are

θ_{1} = - 20^{\circ}

θ_{2} = 0^{\circ}

and

θ_{3} = 20^{\circ}

, respectively. They can be regarded as successfully detected when all of their absolute DOA estimation errors are within

{0.1}^{\circ}

. It can be observed that when SNR is high enough, all methods provide

100 %

target resolution probability. Owing to the increased number of virtual array elements, more diversity gain is obtained in the proposed algorithm. Consequently, the proposed method provides higher resolution probability than the ESPRIT-like,

l_{1}

-SVD and

l_{1}

-SRDML algorithms in all SNR regions, as shown in the simulation results of Figure 7.

6. Conclusions

In this paper, we have proposed the noncircular source-based sparse representation algorithm for DOA estimation in MIMO radar with mutual coupling. On the basis of the error matrix structure and the properties of noncircular signals, the proposed method eliminates the effect of mutual coupling and obtains the extended observation matrix. Then, the joint reweighted sparse representation scheme is constructed to achieve the DOA estimation. The computational complexity of the proposed method has been analyzed, and the simulation results have verified that in the presence of mutual coupling, the proposed algorithm provides higher resolution and better angle estimation performance than the ESPRIT-like,

l_{1}

-SVD and

l_{1}

-SRDML algorithms.

Acknowledgments

This work was supported by the China Scholarship Council and the Fundamental Research Funds for the Central Universities (HEUCF160401).

Author Contributions

Weidong Zhou and Jing Liu contributed to the new method and completed this paper. Pengxiang Zhu, Wenhe Gong and Jiaxin Hou assisted in performing the simulation experiments and analyzing the results.

Conflicts of Interest

The authors declare no conflict of interest.

References

Fishler, E.; Haimovich, A.; Blum, R. Performance of MIMO radar systems: Advantages of angular diversity. In Proceedings of the Thirty-Eighth Asilomar Conference on Signals, Systems and Computers, Pacific Grove, CA, USA, 7–10 November 2004; pp. 305–309.
Wang, X.; Jeske, D.; Serpedin, E. An overview of a class of clock synchronization algorithms for wireless sensor networks: A statistical signal processing perspective. Algorithms 2015, 8, 590–620. [Google Scholar] [CrossRef]
Gao, X.; Zhang, X.F.; Feng, G.P.; Wang, Z.Q.; Xu, D.Z. On the MUSIC-derived aprroaches of angle estimation for bistatic MIMO radar. In Proceedings of the 2009 International Conference on Wireless Networks and Information Systems, TBD, Shanghai, China, 28–29 December 2009; pp. 343–346.
Chen, D.; Chen, B.; Qin, G. Angle estimation using ESPRIT in MIMO radar. Electron. Lett. 2008, 44, 770–771. [Google Scholar]
Zhang, X.; Huang, Y.; Chen, C.; Li, J.; Xu, D. Reduced-complexity Capon for derection of arrival estimation in a monostatic mutiple-input multiple-output radar. IET Radar Sonar Navig. 2012, 8, 796–801. [Google Scholar] [CrossRef]
He, J.; Gao, M.W.; Zhang, L.; Wu, H. Sparse signal recovery from fixed low-rank subspace via compressive measurement. Algorithms 2013, 6, 871–882. [Google Scholar] [CrossRef]
Burns, B.L.; Wilson, N.E.; Thomas, M.A. Group sparse reconstruction of multi-dimensional spectroscopic imaging in human brain in vivo. Algorithms 2014, 7, 276–294. [Google Scholar] [CrossRef]
Liu, Y.J.; Tao, T.Y. A CS recovery algorithm for model and time delay identification of MISO-FIR systems. Algorithms 2015, 8, 743–753. [Google Scholar] [CrossRef]
Malioutov, D.; Cetin, M.; Willsky, A.S. A sparse signal reconstruction perspective for source localization with sensor arrays. IEEE Trans. Signal Process. 2005, 53, 3010–3022. [Google Scholar] [CrossRef]
Yin, J.; Chen, T. Direction-of-arrival estimation using a sparse representation of array covariance vectors. IEEE Trans. Signal Process. 2011, 59, 4489–4493. [Google Scholar] [CrossRef]
Wang, X.P.; Wang, W.; Li, X.; Liu, J. Real-valued covariance vector sparsity-inducing DOA estimation for monostatic MIMO radar. Sensors 2015, 15, 28271–28286. [Google Scholar] [CrossRef] [PubMed]
Dai, J.; Xu, X.; Zhao, D. Direction-of-arrival estimation via real-valued sparse representation. IEEE Antennas Wirel. Propag. Lett. 2013, 12, 376–379. [Google Scholar] [CrossRef]
Liu, X.; Liao, G. Direction finding and mutual coupling estimation for bistatic MIMO radar. Signal Process. 2012, 92, 517–522. [Google Scholar] [CrossRef]
Zheng, Z.D.; Zhang, J.; Zhang, J.Y. Joint DOD and DOA estimation of bistatic MIMO radar in the presence of unknown mutual coupling. Signal Process. 2012, 92, 3039–3048. [Google Scholar] [CrossRef]
Dai, J.; Zhao, D.; Ji, X. A sparse representation method for DOA estimation with unknown mutual coupling. IEEE Antennas Wirel. Propag. Lett. 2012, 11, 1210–1213. [Google Scholar] [CrossRef]
Liu, J.; Wang, X.P.; Zhou, W.D. Covariance vector sparsity-aware DOA estimation for monostatic MIMO radar with unknown mutual coupling. Signal Process. 2016, 119, 21–27. [Google Scholar] [CrossRef]
Bencheikh, M.L.; Wang, Y. Non circular ESPRIT-RootMUSIC joint DOA-DOD estimation in bistatic MIMO radar. In Proceedings of the 2011 7th International Workshop on Systems, Signal Processing and Their Applications (WOSSPA), Tipaza, Algeria, 9–11 May 2011; pp. 51–54.
Picinbono, B. On circularity. IEEE Trans. Signal Process. 1994, 42, 3473–3482. [Google Scholar] [CrossRef]
Yang, M.L.; Chen, B.X.; Yang, X.Y. Conjugate ESPRIT algorithm for bistatic MIMO radar. Electron. Lett. 2010, 46, 1692–1694. [Google Scholar] [CrossRef]
Wang, W.; Wang, X.; Song, H.; Ma, Y. Conjugate ESPRIT for DOA estimation in monostatic MIMO radar. Signal Process. 2013, 93, 2070–2075. [Google Scholar] [CrossRef]
Li, J.; Stoica, P. MIMO radar with colocated antennas. IEEE Signal Process. Mag. 2007, 24, 106–114. [Google Scholar] [CrossRef]
Wang, W.J.; Ren, S.W.; Ding, Y.T.; Wang, H.Y. An efficient algorithm for direction finding against unknown mutual coupling. Sensors 2014, 14, 20064–20077. [Google Scholar] [CrossRef] [PubMed]
Wang, Y.X.; Trinkle, M.; Ng, B.W.-H. DOA estimation under unknown mutual coupling and multipath with improved effective array aperture. Sensors 2015, 15, 30856–30869. [Google Scholar] [CrossRef] [PubMed]
Friedlander, B.; Weiss, A.J. Direction finding in the presence of mutual coupling. IEEE Trans. Antenn. Propag. 1991, 39, 273–284. [Google Scholar] [CrossRef]
Haardt, M.; Romer, F. Enhancements of unitary ESPRIT for non-circular sources. In Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing, Montreal, QC, Canada, 17–21 May 2004; pp. 1520–6149.
Abeida, H.; Delmas, J.P. MUSIC-like estimation of direction of arrival for noncircular sour. IEEE Trans. Signal Process. 2006, 54, 2678–2690. [Google Scholar] [CrossRef]
Sturm, J.F. Using SeDuMi, A Matlab toolbox for optimization ove symmetric cones. Optim. Method Softw. 2009, 11–12, 625–653. [Google Scholar]
Grant, M.; Boyd, S. CVX: MATLAB Software for Disciplined Convex Programming, Version 2.1; CVX Research: Austin, TX, USA, 2012.
Taboga, M. Lectures on Probability Theory and Mathematical Statistics; Amazon CreateSpace: North Charleston, SC, USA, 2012; pp. 469–473. [Google Scholar]
Wax, M.; Kailath, T. Detection of signals by information theoretic criteria. IEEE Trans. Acoust. Speech Signal Process. 1985, 33, 387–392. [Google Scholar] [CrossRef]

Figure 1. Configuration of MIMO radar in the presence of mutual coupling.

Figure 2. Difference between monostatic MIMO radar and bistatic MIMO radar.

Figure 3. Spatial spectrum of the proposed method with target number

P = 3

and

P = 5

Figure 3. Spatial spectrum of the proposed method with target number

P = 3

and

P = 5

Figure 4. RMSE versus SNR in ESPRIT-like,

l_{1}

-SVD and

l_{1}

-SRDML (sparse representation deterministic maximum likelihood) methods for nonzero mutual coupling coefficient number

K = 2

Figure 4. RMSE versus SNR in ESPRIT-like,

l_{1}

-SVD and

l_{1}

-SRDML (sparse representation deterministic maximum likelihood) methods for nonzero mutual coupling coefficient number

K = 2

Figure 5. RMSE versus SNR in the ESPRIT-like,

l_{1}

-SVD and

l_{1}

-SRDML methods for nonzero mutual coupling coefficient number

K = 3

Figure 5. RMSE versus SNR in the ESPRIT-like,

l_{1}

-SVD and

l_{1}

-SRDML methods for nonzero mutual coupling coefficient number

K = 3

Figure 6. RMSE versus snapshots in the ESPRIT-like,

l_{1}

-SVD and

l_{1}

-SRDML methods when SNR = 0 dB.

Figure 6. RMSE versus snapshots in the ESPRIT-like,

l_{1}

-SVD and

l_{1}

-SRDML methods when SNR = 0 dB.

Figure 7. Target resolution probability versus SNR in the ESPRIT-like,

l_{1}

-SVD and

l_{1}

-SRDML methods with snapshot number

J = 100

Figure 7. Target resolution probability versus SNR in the ESPRIT-like,

l_{1}

-SVD and

l_{1}

-SRDML methods with snapshot number

J = 100

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).

Share and Cite

MDPI and ACS Style

Zhou, W.; Liu, J.; Zhu, P.; Gong, W.; Hou, J. Noncircular Sources-Based Sparse Representation Algorithm for Direction of Arrival Estimation in MIMO Radar with Mutual Coupling. Algorithms 2016, 9, 61. https://doi.org/10.3390/a9030061

AMA Style

Zhou W, Liu J, Zhu P, Gong W, Hou J. Noncircular Sources-Based Sparse Representation Algorithm for Direction of Arrival Estimation in MIMO Radar with Mutual Coupling. Algorithms. 2016; 9(3):61. https://doi.org/10.3390/a9030061

Chicago/Turabian Style

Zhou, Weidong, Jing Liu, Pengxiang Zhu, Wenhe Gong, and Jiaxin Hou. 2016. "Noncircular Sources-Based Sparse Representation Algorithm for Direction of Arrival Estimation in MIMO Radar with Mutual Coupling" Algorithms 9, no. 3: 61. https://doi.org/10.3390/a9030061

APA Style

Zhou, W., Liu, J., Zhu, P., Gong, W., & Hou, J. (2016). Noncircular Sources-Based Sparse Representation Algorithm for Direction of Arrival Estimation in MIMO Radar with Mutual Coupling. Algorithms, 9(3), 61. https://doi.org/10.3390/a9030061

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Menu

Noncircular Sources-Based Sparse Representation Algorithm for Direction of Arrival Estimation in MIMO Radar with Mutual Coupling

Abstract

1. Introduction

2. Problem Formulation

2.1. MIMO Radar System Model with Mutual Coupling

2.2. Noncircular Signals

3. The Proposed Algorithm

3.1. Mutual Coupling Elimination

3.2. Noncircular Signal-Based Extended Matrix Construction

3.3. Joint Reweighted Sparse Representation-Based DOA Estimation Scheme

4. Related Remarks

5. Simulation Results

6. Conclusions

Acknowledgments

Author Contributions

Conflicts of Interest

References

Share and Cite

Article Metrics

Article Access Statistics

Further Information

Guidelines

MDPI Initiatives

Follow MDPI