 Research
 Open Access
 Published:
Sparse reconstruction for directionofarrival estimation using multifrequency coprime arrays
EURASIP Journal on Advances in Signal Processing volume 2014, Article number: 168 (2014)
Abstract
In this paper, multifrequency coprime arrays are employed to perform directionofarrival (DOA) estimation with enhanced degrees of freedom (DOFs). Operation at multiple frequencies creates additional virtual elements in the difference coarray of the coprime array corresponding to the reference frequency. Sparse reconstruction is then used to fully exploit the enhanced DOFs offered by the multifrequency coarray, thereby increasing the number of resolvable sources. For the case where the sources have proportional spectra, the received signal vectors at the different frequencies are combined to form an equivalent single measurement vector model corresponding to the multifrequency coarray. When the sources have nonproportional spectra, a group sparsitybased reconstruction approach is used to determine the direction of signal arrivals. Performance evaluation of the proposed multifrequency approach is performed using numerical simulations for both cases of proportional and nonproportional source spectra.
1 Introduction
Directionofarrival (DOA) estimation has been an area of continued research interest due to its wide range of applications in radar, sonar, and wireless communications [1–3]. A main parameter in DOA estimation is the maximum number of sources that can be resolved. Traditional highresolution DOA techniques, such as MUSIC [4] and ESPRIT [5], can only estimate up to (N 1) sources when applied to an N element uniform linear array (ULA). Numerous nonuniform array geometries and signal processing techniques have been introduced to increase the number of resolvable sources beyond that offered by a ULA for a given number of physical sensors [6–13].
Minimum redundancy arrays (MRAs) and minimum hole arrays (MHAs) are two common classes of nonuniform linear arrays [6–9]. Both MRAs and MHAs provide the ability to resolve more sources than the number of physical sensors by reducing the number of redundant virtual elements in the difference coarray. The difference coarray is defined as the set of all pairwise differences of array element locations, and thus, it specifies the set of 'lags’ at which the spatial correlation function may be estimated [10, 11]. For a given number of physical sensors, MRAs are arrays with the lowest possible redundancy and without any missing lags or 'holes’ in the corresponding coarray, whereas MHAs (also known as Golomb arrays) have a minimum number of holes and zero redundancy. More recently, nested coprime array configurations have been proposed [12, 13]. Nested structure is obtained by systematically nesting two uniform linear subarrays, with one subarray assuming a unit interelement spacing, and can provide O(N^{2}) degrees of freedom (DOFs)_ using only N physical sensors. A nested array generates a coarray with no holes. The coprime array consists of two uniform linear subarrays having M and N sensors with specific interelement spacings, where M and N are coprime, and offers O(MN) degrees of freedom. The coprime array produces a coarray that has both redundancy and holes.
Highresolution DOA estimation with nonuniform arrays can be accomplished based on two main approaches, namely, covariance matrix augmentation [14] and spatial smoothing method for covariance matrix construction [12, 13, 15]. In the former approach, the constructed augmented covariance matrix is not guaranteed to be positive semidefinite and requires positive definite Toeplitz completion [16, 17]. The latter approach vectorizes the covariance matrix of the nonuniform array to emulate observations at a virtual array whose sensor positions are given by the difference coarray. Since the source signals are replaced by their powers in this case, rendering the source environment coherent, spatial smoothing is utilized to decorrelate the sources and to restore the rank of the corresponding covariance matrix. However, the spatial smoothingbased method can only be applied to that part of the difference coarray that has a contiguous number of elements without any holes, implying that this technique cannot fully utilize the DOFs offered by coprime arrays.
In [18], multiple frequencies were employed to exploit all of the DOFs of coprime arrays, thus, increasing the number of resolvable sources. Measurements made at carefully chosen additional frequencies were used to fill in the missing elements in the difference coarray [19]. In doing so, the filled part of the difference coarray is extended, which, in turn, increases the maximum number of sources resolved by highresolution DOA estimation techniques. However, the increase in DOFs comes with a restriction on the sources’ spectra. More specifically, the sources are required to have proportional spectra at the considered frequencies [18, 20]. Although this method provides the ability to utilize all of the DOFs of the coprime array, only a small portion of the additional measurements at frequencies other than the reference frequency are used; the rest are discarded.
In this paper, sparse reconstruction is considered to make use of the full measurement set corresponding to the multifrequency operation for DOA estimation with coprime arrays. This enhances the DOFs beyond those offered by a singlefrequency operation due to the additional virtual elements generated in the coarray under multifrequency operation. For sources with proportional spectra, the observations at the different frequencies are cast as a single measurement vector model, which corresponds to a virtual array whose element positions are given by the union set of the difference coarrays corresponding to the multiple operational frequencies. Sparse reconstruction can then be applied for estimating the directions of signal arrivals. For the case where the sources have nonproportional spectra, the source signal vectors corresponding to the different frequencies have a common support, as the sources maintain their DOA even if their power varies with frequency. The common structure property of the sparse source vectors suggests the application of a group sparse reconstruction. It is noted that sparse recovery was previously applied for DOA estimation with coprime arrays in [21, 22]; however, it was limited to a singlefrequency operation and did not consider enhancement of the DOFs of coprime arrays through multifrequency operation.
Performance evaluation of the proposed sparsitybased methods is conducted using numerical simulations. We consider three different cases for DOA estimation using sparse reconstruction at multiple frequencies. In the first case, all sources are assumed to have the same bandwidth and all sensors operate at the same multiple frequencies. The second and third cases violate the above assumption with a subset of sensors only operating at multiple frequencies and the sources having nonidentical bandwidth but overlapping spectra.
The remainder of the paper is organized as follows. In Section 2, the multifrequency signal model for coprime arrays is presented. In Section 3, the sparse reconstructionbased DOA estimation for multifrequency coprime arrays under proportional spectra is discussed. The case of sources with nonproportional spectra is considered in Section 4, and the group sparsitybased reconstruction is presented. The performance of the proposed methods is evaluated in Section 5 through numerical simulations, and Section 6 concludes the paper.
Notation
Vectors and matrices are denoted by lowercase and uppercase bold characters, respectively. Superscript (.)^{T} denotes the transpose of a matrix or a vector, whereas their conjugate transpose is denoted by superscript (.)^{H}. The Kronecker product and the KhatriRao product [1] are denoted by the symbols ⊗ and ⊙, respectively. E{.} denotes the statistical expectation operator, vec(.) denotes the vectorized form of a matrix which is obtained by stacking the columns of the matrix to form a long vector, and bdiag{∙} denotes block diagonal matrix.
2 Multifrequency signal model
In its basic configuration, a coprime array consists of two uniformly spaced linear subarrays. The first subarray has M elements with Nd_{0} interelement spacing, and the second one has N elements with Md_{0} spacing, where M and N are coprime numbers, and d_{0} = λ_{0}/2 is the unit spacing with λ_{0} being the wavelength at a reference frequency ω_{0}[13]. In this work, we deal with an extended coprime array configuration, proposed in [15], which has twice the number of elements in one of the subarrays. More specifically, we assume M to be less than N with the first subarray having 2 M elements, as shown in Figure 1. The elements of the two subarrays are arranged along a single line with the zeroth elements coinciding, resulting in a coprime array with a total of (2 M + N 1) nonuniformly spaced physical elements. The difference coarray of the extended coprime array can be expressed as:
where 0 ≤ n ≤ N 1, and 0 ≤ m ≤2 M 1. The coarray has an aperture of length 2(2 M 1) Nd_{0} with contiguous elements between  (MN + M 1)d_{0} and (MN + M 1)d_{0}, as depicted in Figure 2.
Assume that D sources are impinging on the (2 M + N 1)  element coprime array from directions [θ_{1}, θ_{2},…, θ_{ D }], where θ is the angle relative to broadside. The (2 M + N 1) × 1 received data vector is expressed as:
where s(ω_{0}) = [s_{1}(ω_{0}) s_{2}(ω_{0}) … s_{ D }(ω_{0})]^{T} is the source signal vector at ω_{0}, and n(ω_{0}) is the (2 M + N 1) × 1 noise vector at ω_{0}. The (2 M + N 1) × D matrix A(ω_{0}) is the array manifold at ω_{0}, whose columns are the steering vectors corresponding to the sources directions. That is, A = [a(θ_{1}) a(θ_{2}) … a(θ_{ D })] with the steering vector a(θ_{ d }) corresponding to the direction θ_{ d } given by:
Here, k_{0} = 2π/λ_{0} is the wavenumber at the reference frequency ω_{0}, and x_{ i }, i =0, 1,…, 2 M + N 1, is the location of the i th array element. Assuming that the sources are uncorrelated and the noise is spatially and temporally white with variance {\sigma}_{n}^{2} and uncorrelated from the sources, the autocorrelation matrix of the received data is given by:
where R_{ ss }(ω_{0}) is the source correlation matrix which is diagonal with the source powers at ω_{0}, {\sigma}_{1}^{2}\left({\omega}_{0}\right),\phantom{\rule{0.12em}{0ex}}{\sigma}_{2}^{2}\left({\omega}_{0}\right),\dots ,\phantom{\rule{0.12em}{0ex}}{\sigma}_{D}^{2}\left({\omega}_{0}\right), populating its main diagonal, and I is a (2 M + N 1) × (2 M + N 1) identity matrix. In practice, the autocorrelation matrix is estimated as a sample average of the received signal snapshots.
Following the formulation in [13], the autocorrelation matrix is vectorized as:
where Ã(ω_{0}) = A^{*}(ω_{0}) ⊙ A(ω_{0}) = [a^{*}(θ_{1}) ⊗ a(θ_{1}) … a^{*}(θ_{ D }) ⊗ a(θ_{ D })], p(ω_{0}) is the sources powers vector at ω_{0}, \mathit{p}\left({\omega}_{0}\right)={\left[{\sigma}_{1}^{2}\left({\omega}_{0}\right)\phantom{\rule{0.12em}{0ex}}{\sigma}_{2}^{2}\left({\omega}_{0}\right)\dots \phantom{\rule{0.12em}{0ex}}{\sigma}_{D}^{2}\left({\omega}_{0}\right)\right]}^{T}, and ĩ is the vectorized form of I. The vector z(ω_{0}) behaves as the received signal vector at a longer virtual array with sensor positions given by the difference coarray at ω_{0} of the physical array. In this model, the sources are replaced by their respective powers and, as such, act as mutually coherent sources, and the noise is deterministic. Traditional subspacebased highresolution methods, such as MUSIC, can no longer be applied directly to perform DOA estimation. Spatial smoothing can be used to restore the rank of the correlation matrix of z(ω_{0}) [23]. However, it can only be applied to the filled part of the difference coarray and the usable DOFs are reduced to approximately onehalf of the total number of contiguous coarray elements. Sparsitybased DOA estimation can help extend the usable DOFs to the number of positive lags in the coarray [24].
Consider operating the physical coprime array at Q different frequencies with the q th frequency given by ω_{ q } = α_{ q }ω_{0}, where α_{ q } is a constant. Note that it is not required for the reference frequency ω_{0} to be one of the Q operational frequencies. If it is included in the operational frequency set, the corresponding α_{ q } assumes a unit value. The received signal at each considered frequency can be extracted by decomposing the array output vector into multiple nonoverlapping narrowband components by using the discrete Fourier transform (DFT) [25, 26]. The observation time is assumed to be sufficiently long to resolve the different frequencies. The received signal vector corresponding to the q th operational frequency can be expressed as:
Here, s(ω_{ q }) and n(ω_{ q }) are the source signal and noise vectors at ω_{ q }, and A(ω_{ q }) is the array manifold at ω_{ q } with its (i, j) th element given by:
where k_{ q } is the wave number at ω_{ q }. We observe from (7) that the array manifold at ω_{ q } is equivalent to the array manifold of a scaled version of the physical coprime array, with the position of the i th sensor in this scaled array given by α_{ q }x_{ i }. This, in turn, results in the difference coarray at ω_{ q } to be a scaled version of the difference coarray of Figure 2 (at the reference frequency ω_{0}), with α_{ q } being the scaling factor [19]. If ω_{ q } is higher than the reference frequency, ω_{0}, the coarray at ω_{ q } is an expanded version of the one at the reference frequency. On the other hand, for ω_{ q } lower than ω_{0}, the equivalent coarray at ω_{ q } is a contracted version of that in Figure 2. For illustration, consider an extended coprime array with M = 3 and N = 7 and the sensor positions given by [0d_{0} 3d_{0} 6d_{0} 7d_{0} 9d_{0} 12d_{0} 14d_{0} 15d_{0} 18d_{0} 21d_{0} 28d_{0} 35d_{0}]. The corresponding difference coarray at ω_{0} is shown in Figure 3a. Operating the array at frequency ω_{1} = 8/7ω_{0}, which is larger than ω_{0}, results in stretching the difference coarray of Figure 3a, as shown in Figure 3b. On the other hand, if the array is operated at a smaller frequency, say ω_{1} = 6/7ω_{0}, the difference coarray undergoes contraction as depicted in Figure 3(c).
For multifrequency DOA estimation, we employ the normalized autocorrelation matrices at each of the Q operational frequencies [20]. The (i,j)th element of the normalized autocorrelation matrix {\overline{R}}_{xx}\left({\omega}_{q}\right) is defined as:
where [x(ω_{ q })]_{ i } is the i th element of the received data vector at frequency ω_{ q }. It can be readily shown that in the normalized autocorrelation matrix {\overline{\mathit{R}}}_{xx}\left({\omega}_{q}\right), the source and noise powers are replaced by their normalized values, which can be expressed as:
where {\sigma}_{k}^{2}\left({\omega}_{q}\right) and {\sigma}_{n}^{2}\left({\omega}_{q}\right) denote the respective powers of the k th source and the noise at ω_{ q }. Similar to (5), we can express {\overline{\mathit{R}}}_{xx}\left({\omega}_{q}\right) in vectorized form as:
where \overline{\mathit{p}}\left({\omega}_{q}\right)={\left[{\overline{\sigma}}_{1}^{2}\left({\omega}_{q}\right)\phantom{\rule{0.12em}{0ex}}{\overline{\sigma}}_{2}^{2}\left({\omega}_{q}\right)\dots \phantom{\rule{0.12em}{0ex}}{\overline{\sigma}}_{D}^{2}\left({\omega}_{q}\right)\right]}^{T}. The measurement vector \overline{\mathit{z}}\left({\omega}_{q}\right) emulates observations at the difference coarray corresponding to ω_{ q }.
The measurement vectors \overline{\mathit{z}}\left({\omega}_{q}\right),\phantom{\rule{0.24em}{0ex}}q=1,2,\dots ,Q can be combined to establish an appropriate multifrequency linear model that permits DOA estimation within the sparse reconstruction framework. In the sequel, we distinguish two cases of normalized source spectra. In the first case, we assume the normalized power of each source to be independent of frequency,
whereas the normalized source powers are allowed to vary with frequency in the second case.
3 Sparsitybased DOA estimation under proportional spectra
We discretize the angular region of interest into a finite set of K ≫ D grid points, \left\{{\theta}_{{g}_{1}},\phantom{\rule{0.12em}{0ex}}{\theta}_{{g}_{2}},\phantom{\rule{0.12em}{0ex}}\dots ,\phantom{\rule{0.12em}{0ex}}{\theta}_{{g}_{K}}\right\}, with {\theta}_{{g}_{1}} and {\theta}_{{g}_{K}} being the limits of the search space. The sources are assumed to be located on the grid. Several methods can be used to modify the model in order to deal with offgrid targets [27, 28]. Then, (11) can be rewritten as:
where the columns of the (2 M + N 1)^{2} × K matrix Ã_{ g }(ω_{ q }) are the steering vectors at ω_{ q } corresponding to the defined angles in the grid. The vector \overline{\mathit{x}}\left({\omega}_{q}\right) is a Dsparse vector whose support corresponds to the source directions with the nonzero values equal to the normalized source powers.
For a high signaltonoise ratio (SNR), a sufficient condition for (12) to hold is that the sources must have proportional spectra at the employed frequencies [20]. That is:
This case is applicable, e.g., when the D sources are BPSK or chirplike signals. Under proportional source spectra, the source vector \overline{\mathit{p}}\left({\omega}_{q}\right) is no longer a function of ω_{ q }, i.e., \overline{\mathit{p}}\left({\omega}_{q}\right)=\overline{\mathit{p}}={\left[{\overline{\sigma}}_{1}^{2}\phantom{\rule{0.12em}{0ex}}{\overline{\sigma}}_{2}^{2}\dots \phantom{\rule{0.12em}{0ex}}{\overline{\sigma}}_{D}^{2}\right]}^{T}for all q, which implies that vector \overline{\mathit{x}}\left({\omega}_{q}\right)=\overline{\mathit{x}}\phantom{\rule{0.4em}{0ex}}\mathrm{\text{for}}\phantom{\rule{0.12em}{0ex}}\mathrm{\text{all}}\phantom{\rule{0.4em}{0ex}}q. As such, the measurement vectors \overline{\mathit{z}}\left({\omega}_{q}\right) at the Q operating frequencies can be stacked to form a single Q(2 M + N 1)^{2} × 1 vector:
where {\overline{\mathit{z}}}_{g}={\left[\overline{\mathit{z}}{\left({\omega}_{1}\right)}^{T}\phantom{\rule{0.12em}{0ex}}\overline{\mathit{z}}{\left({\omega}_{2}\right)}^{T}\dots \phantom{\rule{0.12em}{0ex}}\overline{\mathit{z}}{\left({\omega}_{Q}\right)}^{T}\right]}^{T}, {\tilde{\mathit{i}}}_{g}={\left[{\overline{\sigma}}_{n}^{2}\left({\omega}_{1}\right){\tilde{\mathit{i}}}^{T}\phantom{\rule{0.25em}{0ex}}{\overline{\sigma}}_{n}^{2}\left({\omega}_{2}\right){\tilde{\mathit{i}}}^{T}\dots \phantom{\rule{0.25em}{0ex}}{\overline{\sigma}}_{n}^{2}\left({\omega}_{Q}\right){\tilde{\mathit{i}}}^{T}\right]}^{T}, and the dictionary {\tilde{\mathit{B}}}_{g}={\left[{\tilde{\mathit{A}}}_{g}{\left({\omega}_{1}\right)}^{T}\phantom{\rule{0.25em}{0ex}}{\tilde{\mathit{A}}}_{g}{\left({\omega}_{2}\right)}^{T}\dots \phantom{\rule{0.25em}{0ex}}{\tilde{\mathit{A}}}_{g}{\left({\omega}_{Q}\right)}^{T}\right]}^{T}. The measurement vector is equivalent to that of a virtual array, whose element positions are given by the combined difference coarrays at the Q frequencies, i.e.,
where S_{0} is defined in (1). It is noted that in the case of overlapping points in the Q coarrays, an averaged value of the multiple measurements that correspond to the same coarray location can be used. This results in a reduction in the dimensionality of {\overline{\mathit{z}}}_{g}. More specifically, the length of {\overline{\mathit{z}}}_{g} becomes equal to the total number of unique lags in the combined difference coarray, which is given by:
The dictionary matrix and the noise vector would be changed accordingly.
It should be noted that not all the physical sensors must operate at all Q frequencies. Situations may arise due to cost and hardware restrictions that only a few sensors can accommodate a diverse set of frequencies. The overall difference coarray is still the union of coarrays at the individual frequencies. However, the difference coarray at each frequency may no longer be a scaled version of the difference coarray at the reference frequency.
Given the model in (15), DOA estimation proceeds in terms of sparse signal reconstruction by solving the following constrained minimization problem:
where ϵ is a userspecified bound which depends on the noise variance. The constraint \overline{\mathit{x}}\ge 0 forces the search space to be limited to nonnegative values [22]. This is due to the fact that the nonzero elements of \overline{\mathit{x}} correspond to the normalized source powers, which are always positive. This constraint accelerates the convergence of the solution by reducing the search space. Various techniques can be used to solve the constrained minimization problem in (18), examples being lasso, OMP, and CoSaMP [29–31]. In this paper, we use lasso which solves an equivalent problem to (18):
where the l_{2}norm is the least squares cost function and the l_{1}norm encourages a sparse solution. The regularization parameter λ_{ t } is used to control the weight of the sparsity constraint in the overall cost function. Increasing λ_{ t } results in a sparser solution at the cost of an increased least squares error. Several methods have been proposed to estimate the regularization parameter, such as the discrepancy principle [28, 32] and cross validation [29].
The maximum number of resolvable sources using the proposed method depends on the number of unique lags in the combined difference coarray. According to [33], the sparsitybased minimization problem in (19) is guaranteed to have a unique solution under the condition m ≥2D, where m is equal to the number of independent observations or the number of unique lags in the combined difference coarray. As a result, the maximum number of resolvable sources is equal to the number of unique positive lags in the combined coarray. At the reference frequency, the difference coarray extends from  (2 M 1) Nd_{0} to (2 M 1) Nd_{0}, and it has a total of (M 1) (N 1) holes, which means that the number of unique lags at each frequency is equal to (3MN + M  N), and the highest number of possible unique positive lags is (3MN + M  N 1)/2. Therefore, the maximum number of resolvable sources at each frequency is (3MN + M  N 1)/2. Taking into account the overlap between the lags at the different employed frequencies, the maximum number of resolvable sources with the multifrequency technique is bounded as follows:
The term (Q 1) is subtracted from the upper bound due to the unavoidable overlap between the Q difference coarrays for the zero lag.
4 Sparsitybased DOA estimation under nonproportional spectra
When the source powers vary with frequency, the single measurement vector model of (15) is no longer applicable. However, the D sources have the same directions [θ_{1}, θ_{2},…, θ_{ D }] regardless of their power distribution with frequency. As such, the vectors \overline{\mathit{x}}\left({\omega}_{q}\right),\phantom{\rule{0.12em}{0ex}}q=1,\phantom{\rule{0.12em}{0ex}}2,\phantom{\rule{0.12em}{0ex}}\dots ,Q, in (13) have a common support. That is, if a certain element in, e.g., \overline{\mathit{x}}\left({\omega}_{1}\right) has a nonzero value, the corresponding elements in \overline{\mathit{x}}\left({\omega}_{q}\right),\phantom{\rule{0.24em}{0ex}}q=2,\dots Q, should be also nonzero. The common structure property suggests the application of a group sparse reconstruction. We, therefore, propose a DOA estimation approach based on group sparsity for the nonproportional spectra case.
The received signal vectors \overline{\mathit{z}}\left({\omega}_{q}\right) in (13) corresponding to the Q frequencies are stacked to form a long vector:
where {\tilde{\mathit{C}}}_{g}=\mathrm{\text{bdiag}}\left\{{\tilde{\mathit{A}}}_{g}\left({\omega}_{1}\right),{\tilde{\mathit{A}}}_{g}\left({\omega}_{2}\right),\cdots ,{\tilde{\mathit{A}}}_{g}\left({\omega}_{Q}\right)\right\}, and \stackrel{\u2323}{\mathit{x}}={\left[\overline{\mathit{x}}{\left({\omega}_{1}\right)}^{T}\phantom{\rule{0.12em}{0ex}}\overline{\mathit{x}}{\left({\omega}_{2}\right)}^{T}\dots \phantom{\rule{0.12em}{0ex}}\overline{\mathit{x}}{\left({\omega}_{Q}\right)}^{T}\right]}^{T}. The vector \stackrel{\u2323}{\mathit{x}} is a group sparse vector where each group consists of the source powers corresponding to a specific direction at all operating frequencies. The group sparse solution is obtained by minimizing the following mixed l_{1}  l_{2} norm cost function:
where:
This means that the variables belonging to the same group are combined using the l_{2} norm, and the l_{1} norm is then used across the groups to enforce group sparsity. Different algorithms can be utilized to perform sparse reconstruction with grouped variables. These algorithms include group lasso and block orthogonal matching pursuit (BOMP) [34, 35], among many others. In this paper, group lasso is considered to perform DOA estimation in the case of sources with nonproportional spectra. Further, similar to the method discussed in Section 3, a constraint can be added to force the elements of the solution vector \stackrel{\u2323}{\mathit{x}} to be nonnegative.
It is noted that this formulation results in a smaller number of achievable DOFs compared to the case where the sources have proportional spectra. The maximum number of resolvable sources is now limited by the number of observations or unique lags at each frequency [28]. This means that up to (3MN + M  N 1) sources can be resolved.
5 Numerical results
In this section, we present the DOA estimation results for the proposed sparse reconstruction techniques using multifrequency coprime arrays and provide performance comparison with the sparsitybased approach for singlefrequency coprime array. We consider both cases of proportional and nonproportional source spectra. For all of the examples in this section, an extended coprime array configuration with six physical elements is considered with M and N chosen to be 2 and 3, respectively. The six sensor positions are given by [0, 2d_{0}, 3d_{0}, 4d_{0}, 6d_{0}, 9d_{0}]. The corresponding difference coarray, shown in Figure 4, consists of 17 virtual elements. The coarray aperture extends from 9d_{0} to 9d_{0} with two holes at ± 8d_{0}. Further, in all the examples, the choice of the simulation parameters, such as the SNR and the number of snapshots, is typical of radio frequency (RF) applications.
In the first example, sparse signal reconstruction is applied under single frequency operation to perform DOA estimation. Since the difference coarray has eight positive lags, sparse reconstruction can be applied to resolve up to eight sources. A total of eight BPSK sources, uniformly spaced between 60° and 60°, are considered. The number of snapshots used is 1,000. Spatially and temporally white Gaussian noise is added to the observations, and the SNR is set to 10 dB for all sources. The search space is discretized uniformly between 90° and 90° with a 0.2° step size, and the regularization parameter λ_{ t }, is empirically chosen as 0.7 in this example. The normalized spectrum obtained using sparse signal recovery is shown in Figure 5. The dashed vertical lines in the figure indicate the true source directions. A small bias can be noticed in the estimates, and the root mean squared error (RMSE), computed across the angles of arrival, is found to be 1.05° in this case.
In the second example, sparse reconstruction is applied under dualfrequency operation. The physical coprime array is now operated at both frequencies ω_{0} and ω_{1} = 8/9ω_{0}. Sources with proportional spectra are assumed, and thus, the single measurement vector formulation of Section 3 can be used. The combined difference coarray is shown in Figure 6. It has a total of 33 unique lags, which makes it capable of resolving up to 16 sources, theoretically. However, this number is not achievable because of the high mutual coherence of the dictionary. Since some of the virtual sensors in the combined coarray are closely separated, leading to highly correlated observations, deterioration in performance is observed if the number of sources is increased beyond 11. We consider 11 BPSK sources with proportional spectra, which are uniformly spaced between 75° and 75°. The SNR is set to 10 dB for the sources at the two frequencies, and the total number of snapshots is equal to 2,000. The regularization parameter λ_{ t } is set to 0.25, and the search space is divided into 181 bins of size 1°. Figure 7 shows the normalized spectrum obtained using this method. It is evident that all the sources are correctly resolved. The RMSE in this example is equal to 0.84°. A different choice of the two operational frequencies may reduce the mutual coherence, thereby permitting a larger number of sources to be estimated. For illustration, the second frequency is now set to ω_{1} = 2ω_{0}. By choosing a frequency which is an integer multiple of ω_{0}, the combined coarray positions are guaranteed to be integer multiples of d_{0}. As a result, the minimum separation between two consecutive coarray elements is equal to d_{0}. The combined difference coarray is shown in Figure 8. The coarray has 13 unique positive lags, which means that the maximum number of resolvable sources is equal to 13. This is tested by considering 13 uniformly spaced sources between 75° to 75°. The SNR is again set to 10 dB, and the number of snapshots is set to 2,000. The regularization parameter is again set to 0.25, and the search space is divided into 181 angle bins. Figure 9 shows the normalized spectrum using the dualfrequency sparse reconstruction method. It is evident that all the sources are correctly estimated. The RMSE is found to be 0.26° in this case.
In the following example, the entire array is operated at ω_{0}, but only the elements at [2d_{0} 4d_{0} 9d_{0}] also operate at the second frequency ω_{1} = 2ω_{0}. The combined difference coarray is shown in Figure 10, where the difference coarray at ω_{0} is shown in black; and the additional lags, obtained by operating the subarray at ω_{1}, are shown in red. The overall difference coarray has ten positive lags which imply that up to ten sources can be resolved. This is tested by considering ten sources with respective DOAs [60°, 49°, 29°, 20°, 9°, 3°, 18°, 29°, 47°, 60°]. The number of snapshots is set to 2,000 at each frequency, and the SNR is 10 dB. The regularization parameter is set to 0.7 in this example, and the search space is kept the same. Figure 11 shows the normalized spectrum using the dualfrequency sparse reconstruction method. It can be noticed that all the sources are correctly estimated, and the corresponding RMSE is 0.86°.
The following example examines the case when sources have nonproportional spectra. In this case, group sparse reconstruction is applied. The two operational frequencies are selected to be ω_{0} and 2ω_{0}. Eight sources with nonproportional spectra are considered. The SNR of all the sources at the first frequency is set to 10 dB. At the second frequency, the SNR of each source is a realization of a uniformly distributed random variable between 5 and 15 dB. This ensures that the sources have nonproportional spectra. The noise variance is set to unity at the two frequencies, and a total of 2,000 snapshots are used. Figure 12 shows the normalized spectrum obtained using the formulation in Section 3 which mistakenly assumes proportional source spectra. Consequently, this method is expected to fail as evident in the spectrum of Figure 12. One of the sources is not resolved and several spurious peaks appear in the spectrum. The DOA estimation is next repeated using group sparse reconstruction which was discussed in Section 4. This method does not require the sources to have proportional spectra. The mean of the recovered spectra at the two employed frequencies is computed and shown in Figure 13. It can be seen that group sparse reconstruction is successful in localizing the DOAs of all the sources. The RMSE is found to be 0.6° in this case.
The next example confirms the increase in the number of resolvable sources by using group sparse reconstruction compared to the singlefrequency sparse reconstruction. As stated in the first example, the maximum number of resolvable sources using singlefrequency sparse reconstruction is equal to the number of unique positive lags in the difference coarray, which is eight in this case. A total of 16 sources with nonproportional spectra is considered in the example. The sources are uniformly spaced between 75° and 75°. Twenty uniformly spaced frequencies between ω_{0} and 2ω_{0} are employed. The SNR of each source at each frequency is chosen randomly between 5 and 5 dB, and the number of snapshots at each frequency is set to 1,000. Figure 14 shows the normalized mean spectrum obtained using group sparse reconstruction. It can be seen that all the sources are correctly estimated, and the RMSE is equal to 0.35° in this case. Figure 15 shows the normalized spectrum for the singlefrequency sparse reconstruction case. This figure confirms that sparse reconstruction using a single frequency completely fails in estimating the sources. This is due to the fact that singlefrequency sparse reconstruction can only resolve up to eight sources which is smaller than the total number of sources in this example.
The final example examines the case where the source signals have overlapping spectra but do not share the same bandwidth. Group sparse reconstruction can still be used to perform DOA estimation. Thirty percent of the source powers at the employed frequencies in the previous example are randomly set to zero. The remaining parameters are kept the same. Figure 16 shows the normalized spectrum using group sparse reconstruction. It is evident that all sources are correctly estimated. Some spurious peaks are present in the spectrum, and an increase in the estimates bias is obtained. The RMSE is found to be 0.61°.
6 Conclusions
A sparse reconstruction method has been proposed for DOA estimation using multifrequency coprime arrays. The proposed approach offers an enhancement in the degrees of freedom over the singlefrequency coprime array. For sources with proportional spectra, all observations at the employed frequencies are combined to form a received signal vector at a larger virtual array, whose elements are given by the combination of the difference coarrays at the individual frequencies, thereby increasing the number of resolvable sources. In the case of sources with nonproportional spectra, the common support that is shared by the observations at the employed frequencies is exploited through group sparse reconstruction. Although the offered degrees of freedom are less than those of the multifrequency approach for proportional spectra, they exceed those offered by singlefrequency coprime array with sparse reconstruction. Numerical examples demonstrated the superior performance of the proposed multifrequency approach compared to its singlefrequency counterpart.
Abbreviations
 CoSaMP:

Compressive sampling matched pursuit
 DFT:

Discrete fourier transform
 DOA:

Direction of arrival
 DOF:

Degree of freedom
 Lasso:

Least absolute shrinkage and selection operator
 MHA:

Minimum hole array
 MRA:

Minimum redundancy array
 OMP:

Orthogonal matching pursuit
 RMSE:

Root mean squared error
 SNR:

Signaltonoise ratio
 ULA:

Uniform linear array.
References
Van Trees HL: Optimum Array Processing: Part IV of Detection, Estimation, and Modulation Theory. John Wiley and Sons, New York; 2002.
Chandran S: Advances in DirectionofArrival Estimation. Artech House, Norwood, MA; 2006.
Tuncer TE, Friedlander B: Classical and Modern DirectionofArrival Estimation. Academic Press (Elsevier), Boston, MA; 2009.
Schmidt R: Multiple emitter location and signal parameter estimation. IEEE. Trans. Antenn. Propag. 1986, 34(3):276280. 10.1109/TAP.1986.1143830
Roy R, Kailath T: ESPRITEstimation of signal parameters via rotational invariance techniques. IEEE. Trans. Acoust. Speech. Signal. Process. 1989, 37(7):984995. 10.1109/29.32276
Moffet A: Minimumredundancy linear arrays. IEEE. Trans. Antenn. Propag. 1968, 16(2):172175. 10.1109/TAP.1968.1139138
Bloom GS, Golomb SW: Application of numbered undirected graphs. Proc. IEEE 1977, 65(4):562570.
Chambers C, Tozer TC, Sharman KC, Durrani TS: Temporal and spatial sampling influence on the estimates of superimposed narrowband signals: when less can mean more. IEEE Trans. Signal Process. 1996, 44(12):30853098. 10.1109/78.553482
Ma WK, Hsieh TH, Chi CY DOA estimation of quasistationary signals via KhatriRao subspace. In Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2009. Taipei, Taiwan; 2009:21652168.
Johnson DH, Dudgeon DE: Array Signal Processing: Concepts and Techniques. Prentice Hall, Englewood, NJ; 1993.
Hoctor RT, Kassam SA: The unifying role of the coarray in aperture synthesis for coherent and incoherent imaging. Proc. IEEE 1990, 78(4):735752. 10.1109/5.54811
Pal P, Vaidyanathan PP: Nested arrays: a novel approach to array processing with enhanced degrees of freedom. IEEE Trans. Signal Process. 2010, 58(8):41674181.
Vaidyanathan PP, Pal P: Sparse sensing with coprime samplers and arrays. IEEE Trans. Signal Process. 2011, 59(2):573586.
Pillai SU, BarNess Y, Haber F: A new approach to array geometry for improved spatial spectrum estimation. Proc. IEEE 1985, 73(10):15221524.
Pal P, Vaidyanathan PP Coprime sampling and the MUSIC algorithm. In Proceedings of the Digital Signal Processing Workshop and IEEE Signal Processing Education Workshop (DSP/SPE). Sedona, AZ; 2011:289294.
Abramovich YI, Gray DA, Gorokhov AY, Spencer NK: Positivedefinite Toeplitz completion in DOA estimation for nonuniform linear antenna arrays. I. Fully augmentable arrays. IEEE Trans. Signal Process. 1998, 46(9):24582471. 10.1109/78.709534
Abramovich YI, Spencer NK, Gorokhov AY: Positivedefinite Toeplitz completion in DOA estimation for nonuniform linear antenna arrays. II. Partially augmentable arrays. IEEE Trans. Signal Process. 1999, 47(6):15021521. 10.1109/78.765119
BouDaher E, Jia Y, Ahmad F, Amin M Directionofarrival estimation using multifrequency coprime arrays. In Proceedings of the 22nd European Signal Processing Conference (EUSIPCO), 2014. Lisbon, Portugal; 2014.
Ahmad F, Kassam SA: Performance analysis and array design for wideband beamformers. J. Electron. Imag. 1998, 7(4):825838. 10.1117/1.482668
Moulton JL, Kassam SA Resolving more sources with multifrequency coarrays in highresolution directionofarrival estimation. In Proceedings of the 43rd Annual Conference on Information Sciences and Systems. Baltimore, MD; 2009:772777.
Zhang YD, Amin MG, Himed B Sparsitybased DOA estimation using coprime arrays. In Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2013. Vancouver, Canada; 2013:39673971.
Pal P, Vaidyanathan PP On application of LASSO for sparse support recovery with imperfect correlation awareness. In Proceedings of the 2012 Conference Record of the Forty Six Asilomar Conference on Signals, Systems and Computers (ASILOMAR). Pacific Grove, CA; 2012:958962.
Shan TJ, Wax M, Kailath T: On spatial smoothing for directionofarrival estimation of coherent signals. IEEE. Trans. Acous. Speech. Signal. Process. 1985, 33(4):806811. 10.1109/TASSP.1985.1164649
Hu N, He Z, Xu X, Bao M: DOA estimation for sparse array via sparse signal reconstruction. IEEE. Trans. Aerospace. Electron. Syst. 2013, 49(2):760773.
Wang H, Kaveh M: Coherent signalsubspace processing for the detection and estimation of angles of arrival of multiple wideband sources. IEEE. Trans. Acoust. Speech. Signal. Process. 1985, 33(4):823831. 10.1109/TASSP.1985.1164667
Yoon Y, Kaplan LM, McClellan JH: TOPS: New DOA estimator for wideband signals. IEEE Trans. Signal Process. 2006, 54(6):19771989.
Tan Z, Nehorai A: Sparse direction of arrival estimation using coprime arrays with offgrid targets. IEEE. Signal. Process. Lett. 2014, 21(1):2629.
Malioutov D, Cetin M, Willsky A: Sparse signal reconstruction perspective for source localization with sensor arrays. IEEE Trans. Signal Process. 2005, 53(8):30103022.
Tibshirani R: Regression shrinkage and selection via the Lasso. J. Roy. Stat. Soc Ser. B. 1996, 58(1):267288.
Tropp JA, Gilbert AC: Signal recovery from random measurements via orthogonal matching pursuit. IEEE. Trans. Informat. Theory. 2007, 53(12):46554666.
Needella D, Tropp JA: CoSaMP: iterative signal recovery from incomplete and inaccurate samples. Appl. Comput. Harmonic. Analysis. 2009, 26(3):301321. 10.1016/j.acha.2008.07.002
Morozov VA: On the solution of functional equations by the method of regularization. Soviet. Math. Doklady. 1966, 7: 414417.
Davenport M, Duarte M, Eldar Y, Kutyniok G: Compressed Sensing: Theory and Applications. Cambridge University Press, Cambridge, UK; 2012:1.
Yuan M, Lin Y: Model selection and estimation in regression with grouped variables. J. Roy. Stat. Soc. Ser. B. 2007, 68(1):4967.
Eldar Y, Kuppinger P, Bolcskei H: Blocksparsity signals: uncertainty relations and efficient recovery. IEEE Trans. Signal Process. 2010, 58(6):30423054.
Acknowledgements
This work was supported by the Office of Naval Research (ONR) under grant N000141310061.
Author information
Affiliations
Corresponding author
Additional information
Competing interests
The authors declare that they have no competing interests.
Authors’ original submitted files for images
Below are the links to the authors’ original submitted files for images.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
To view a copy of this licence, visit https://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
BouDaher, E., Ahmad, F. & Amin, M.G. Sparse reconstruction for directionofarrival estimation using multifrequency coprime arrays. EURASIP J. Adv. Signal Process. 2014, 168 (2014). https://doi.org/10.1186/168761802014168
Received:
Accepted:
Published:
DOI: https://doi.org/10.1186/168761802014168
Keywords
 Sparse reconstruction
 Coprime arrays
 DOA estimation
 Group sparsity