Pitch angle-azimuth angle joint estimation method and device suitable for broadband system

By performing fast Fourier transform and decoupled atomic norm minimization methods on broadband signals, efficient estimation of pitch angle and azimuth angle in broadband systems is achieved, and the problem of being unable to process broadband and high-dimensional signals simultaneously in the prior art is solved. , improves estimation accuracy and calculation speed.

CN119024264APending Publication Date: 2024-11-26INST OF ACOUSTICS CHINESE ACAD OF SCI

Patent Information

Application Number
CN202411049175.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-01
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

The prior art cannot perform effective estimation of pitch angle and azimuth angle for both broadband and high-dimensional signals, especially in practical applications, the signals are often high-dimensional broadband, and traditional methods are difficult to directly apply.

Method used

A joint pitch-azimuth angle estimation method suitable for broadband systems is proposed. By performing fast Fourier transform on data received by uniform rectangular arrays, the signal is converted into a multi-frequency form, using the smallest atomic norm for the decoupled The method of transformation obtains the spatial angular frequency, and then converts it into pitch angle and azimuth angle after matching.

Benefits of technology

It realizes efficient DOA estimation in small samples, eliminates the impact of grid mismatch, improves estimation accuracy, is suitable for broadband signals, has faster calculation speed and wider applicability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119024264A_ABST
    Figure CN119024264A_ABST
Patent Text Reader

Abstract

The invention provides a pitch angle-azimuth angle joint estimation method and device suitable for a broadband system, and the method comprises the steps: carrying out the fast Fourier transform of data received by a uniform rectangular array, and converting a signal into a multi-frequency form; using a decoupled atomic norm minimization method to obtain estimated values of the two space angular frequencies; matching the estimated space angular frequency through a matching algorithm; and converting the matched space angular frequency into a pitch angle and an azimuth angle. The method has the advantages that DOA estimation under the condition of small samples can be realized, and the efficiency is higher; the influence of grid mismatch is eliminated; the precision is greatly improved; the calculation speed is high; and the method is more suitable for actual situations and has wider applicability.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application belongs to the field of array signal processing, and specifically relates to a pitch angle-azimuth angle joint estimation method and device suitable for broadband systems. Background Art

[0002] Direction of arrival (DOA) estimation is a hot research topic in the field of array signal processing and is widely used in sonar, radar, positioning and other application fields. Early DOA estimation was mainly based on traditional beamforming technology, which was limited by the array aperture and could not break through the Rayleigh limit. In order to solve this problem, subspace-based DOA estimation algorithms represented by multiple signal classification (MUSIC) and rotational invariance technology (ESPRIT) came into being and are still widely used today. These methods break through the Rayleigh limit by estimating and applying the signal subspace, greatly improving the resolution capability. However, these methods require signal independence, are sensitive to signal coherence, and require a large amount of snapshot data to obtain a reliable estimate of the signal subspace.

[0003] In recent years, the development of compressed sensing (CS) theory has provided a new solution for DOA estimation. By making full use of the highly sparse characteristics of signals in the DOA domain, it is possible to estimate DOA using a small number of snapshots or even a single snapshot. By dividing the DOA domain into dense grids, the steering vectors corresponding to each grid point can form an overcomplete dictionary, converting the DOA estimation problem into a sparse recovery problem. When the dictionary satisfies the null space property (NSP) or the finite isometry property (RIP), this NP-hard l0 optimization problem can be converted into an l1 optimization problem that can be planned and solved through convex relaxation. When the actual signal does not fall on the divided grid points, it is called grid mismatch. Grid mismatch will greatly reduce the accuracy of the estimation, and when used for positioning, there is also a risk of missing the target. To this end, a series of off-grid algorithms have been proposed. Such algorithms improve the impact of grid mismatch by interpolating between grids, but the problem has not been completely solved because there are still grids.

[0004] The above methods are collectively referred to as grid-based algorithms because they assume the existence of a grid. Compared with grid-based algorithms, the gridless algorithms proposed in recent years do not assume the existence of a grid and directly process continuous DOA. Its representative algorithm is the atomic norm minimization (ANM) method, which can obtain an accurate estimate of DOA by solving the atomic norm minimization problem.

[0005] The traditional ANM algorithm is designed based on one-dimensional narrowband signals. In practical applications, signals are often high-dimensional and wideband. This makes it difficult for the traditional one-dimensional narrowband algorithm to be directly applied in practice.

[0006] In recent years, a lot of research has been done on broadband and high-dimensional signals. For high-dimensional signals, they can be incorporated into the ANM framework by using high-order Toeplitz matrices; for broadband signals, they can be modeled using multi-frequency models to use ANM to estimate their DOA. However, there are few studies on ANM algorithms for broadband high-dimensional signals. Summary of the invention

[0007] The purpose of the present application is to overcome the defect that the prior art cannot estimate the elevation angle and azimuth angle for broadband and high-dimensional signals at the same time.

[0008] In order to achieve the above object, the present application proposes a joint estimation method of elevation angle and azimuth angle applicable to a broadband system, comprising:

[0009] Step 1: Perform fast Fourier transform on the data received by the uniform rectangular array to convert the signal into a multi-frequency form;

[0010] Step 2: Use the decoupled atomic norm minimization method to obtain estimates of the two spatial angular frequencies;

[0011] Step 3: Match the estimated spatial angular frequency through a matching algorithm;

[0012] Step 4: Convert the matched spatial angular frequency into elevation angle and azimuth angle.

[0013] As an improvement of the above method, step 2 comprises:

[0014] For the noise-free case, the following atomic norm minimization problem is constructed:

[0015]

[0016] in, represents the received signal without noise; Y represents the received signal containing noise;

[0017]

[0018] Among them, c i represents the amplitude of the i-th signal; r represents the number of far-field broadband plane wave signals; ‖·‖ is the l2 norm, N f Indicates the number of subbands into which a signal is divided; ⊙ indicates the Khatri-Rao product;

[0019]

[0020] a1(f,ω 1i )=[1,exp(j2πfω 1i ),...,exp(j2πfω1i ·(N1-1));a2(f,ω 2i )=[1,exp(j2πfω 2i ),...,exp(j2πfω 2i ·(N2-1));ω i represents the spatial angular frequency vector of the i-th signal, ω 1i Represents ω i Horizontal component; ω 2i Represents ω i The vertical component; N1 represents the number of array elements in the horizontal direction; N2 represents the number of array elements in the vertical direction; f0 represents the sampling frequency / N f ; c represents the wave propagation speed, d represents the array element spacing; θ i represents the pitch angle of the i-th signal; φ i represents the azimuth of the i-th signal; T represents vector transposition;

[0021] Atomic norm Defined as:

[0022]

[0023] Where t is the stretching constant; inf{·} is the infimum; conv(·) is the convex hull; the atomic set Defined as:

[0024]

[0025] Using the decoupled atomic norm minimization method, the following optimization problem is obtained:

[0026]

[0027] in, T2 is Toeplitz matrix of ; Tr(·) means taking the trace of the matrix; represents conjugate transpose;

[0028] definition as follows:

[0029]

[0030] Mapping Operator is defined as follows:

[0031] So that:

[0032]

[0033] N f ×N f A block matrix where each block is The Toeplitz matrix:

[0034]

[0035] K1 is the number of horizontal components of the estimated spatial angular frequency;

[0036] definition

[0037] Depend on Get S(T1), extract ω from S(T1) 1i ;

[0038] in, where e j is a vector whose elements are all 0 except the jth element which is 1;

[0039] By performing Vandermonde decomposition on T2, we can obtain ω 2i .

[0040] As an improvement of the above method, step 2 comprises:

[0041] For the noisy case, the following atomic norm minimization problem is constructed:

[0042]

[0043] in, represents the received noise-free signal; Y represents the received signal containing noise; β represents the noise power; ‖·‖2 represents the 2-norm of the matrix;

[0044]

[0045] Among them, c i represents the amplitude of the i-th signal; r represents the number of far-field broadband plane wave signals; ‖·‖ is the l2 norm, N f Indicates the number of subbands into which a signal is divided; ⊙ indicates the Khatri-Rao product;

[0046]

[0047] a1(f,ω 1i )=[1,exp(j2πfω 1i ),...,exp(j2πfω 1i ·(N1-1));a2(f,ω 2i)=[1,exp(j2πfω 2i ),...,exp(j2πfω 2i ·(N2-1));ω i represents the spatial angular frequency vector of the i-th signal, ω 1i Represents ω i Horizontal component; ω 2i Represents ω i The vertical component; N1 represents the number of array elements in the horizontal direction; N2 represents the number of array elements in the vertical direction; f0 represents the sampling frequency / N f ; c represents the wave propagation speed, d represents the array element spacing; θ i represents the pitch angle of the i-th signal; φ i represents the azimuth of the i-th signal; T represents vector transposition;

[0048] Atomic norm Defined as:

[0049]

[0050] Where t is the stretching constant; inf{·} is the infimum; conv(·) is the convex hull; the atomic set Defined as:

[0051]

[0052] Using the decoupled atomic norm minimization method, the following optimization problem is obtained:

[0053]

[0054] Among them, ∈ is the noise power; T2 is Toeplitz matrix of ; Tr(·) means taking the trace of the matrix; represents conjugate transpose;

[0055] definition as follows:

[0056]

[0057] Mapping Operator is defined as follows:

[0058] So that:

[0059]

[0060] Nf ×N f A block matrix where each block is The Toeplitz matrix:

[0061]

[0062] K1 is the number of horizontal components of the estimated spatial angular frequency;

[0063] definition

[0064] Depend on Get S(T1), extract ω from S(T1) 1i ;

[0065] in, where e j is a vector whose elements are all 0 except the jth element which is 1;

[0066] By performing Vandermonde decomposition on T2, we can obtain ω 2i .

[0067] As an improvement of the above method, step 3 comprises:

[0068] The matching algorithm is:

[0069]

[0070] f=1,...N f

[0071]

[0072] in,‖·‖ 2,1 for l 2,1 norm; vec(·) means converting the matrix into a vector;

[0073]

[0074] K2 is the number of vertical components of the estimated spatial angular frequency; and are the estimated values ​​of the horizontal and vertical components of the spatial angular frequency, respectively;

[0075] To optimize the results, Normalize each column to get set up By extracting The indices corresponding to the largest r elements can be matched.

[0076] As an improvement of the above method, step 4 includes:

[0077] The pitch angle θ in the i-th signal i and azimuth φ i They are obtained by the following formulas:

[0078]

[0079] The present application also provides a pitch angle-azimuth angle joint estimation device applicable to a broadband system, which is implemented based on the above method, and the device includes:

[0080] The Fourier transform module is used to perform fast Fourier transform on the data received by the uniform rectangular array and convert the signal into a multi-frequency form;

[0081] An estimated spatial angular frequency module, for obtaining estimated values ​​of two spatial angular frequencies using a decoupled atomic norm minimization method;

[0082] A spatial angular frequency matching module, used for matching the estimated spatial angular frequency through a matching algorithm;

[0083] The module for calculating the pitch angle and azimuth angle is used to convert the matched spatial angular frequency into the pitch angle and azimuth angle.

[0084] The present application also provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements any of the above methods when executing the computer program.

[0085] The present application also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor executes any of the methods described above.

[0086] Compared with the prior art, the advantages of this application are:

[0087] 1. Compared with the traditional two-dimensional DOA estimation method based on the subspace method, the method proposed in the present invention can realize DOA estimation in the case of small samples with higher efficiency;

[0088] 2. Compared with the traditional grid-based algorithm, the method proposed in this invention realizes gridless estimation and eliminates the influence of grid mismatch;

[0089] 3. Compared with the traditional narrowband two-dimensional DOA estimation algorithm based on ANM, the method proposed in this invention realizes broadband DOA estimation, fully utilizes the acquired information, and greatly improves the accuracy;

[0090] 4. Compared with the existing wideband one-dimensional DOA estimation algorithm based on ANM, the method proposed in the present invention expands the dimension to two dimensions, which is more in line with the actual situation and has wider applicability;

[0091] 5. Compared with the existing 2DMFANM method and other traditional methods, the calculation speed is faster. BRIEF DESCRIPTION OF THE DRAWINGS

[0092] Figure 1 The figure shows a schematic diagram of a uniform rectangular array (URA) receiving a broadband far-field plane wave signal;

[0093] Figure 2 Shown is a flow chart of a broadband elevation angle-azimuth angle joint estimation method based on an atomic norm minimization method;

[0094] FIG3( a ) shows one of the comparisons of the estimation performance of the 2DMFANM method and the traditional method for the pitch angle and azimuth angle;

[0095] FIG3( b ) shows the second comparison of the estimation performance of the 2DM FANM method and the traditional method for the pitch angle and azimuth angle;

[0096] FIG3(c) shows the third comparison of the estimation performance of the 2DMFANM method and the traditional method for the pitch angle and azimuth angle;

[0097] FIG3( d ) shows one of the performance comparisons of the 2DM FANM method and the traditional method for estimating spatial angular frequency;

[0098] FIG3(e) shows the second comparison of the spatial angular frequency estimation performance between the 2DMFANM method and the traditional method;

[0099] FIG3( f ) shows the third comparison of the spatial angular frequency estimation performance between the 2DM FANM method and the traditional method;

[0100] FIG4( a ) shows a curve comparison of the root mean square error of the spatial angular frequency estimated by the 2DMFANM method and the traditional method as the SNR changes;

[0101] Figure 4(b) shows the curve comparison of the root mean square error of the elevation angle and azimuth angle estimated by the 2DMFANM method and the traditional method as the SNR changes;

[0102] Figure 4(c) shows the curve comparison of the higher estimation times of the 2DMFANM method and the traditional method with the change of SNR;

[0103] Figure 5(a) shows the estimation results of the elevation angle and azimuth angle using the 2DMFANM method;

[0104] Figure 5(b) shows the estimation results of the pitch angle and azimuth angle by the 2DMFDANM method;

[0105] Figure 5(c) shows the estimation result of spatial angular frequency by 2DMFANM method;

[0106] Figure 5(d) shows the estimation result of spatial angular frequency using the 2DMFDANM method. DETAILED DESCRIPTION

[0107] The technical solution of the present application is described in detail below with reference to the accompanying drawings.

[0108] The present invention proposes a pitch angle-azimuth joint estimation method (2DMFDANM) and a device suitable for broadband systems. The method is optimized based on a broadband pitch angle-azimuth joint estimation method (2DMFANM) based on an atomic norm minimization method. The 2DMFANM method uses a multi-frequency model to model broadband signals and proposes a corresponding atomic set. The pitch angle and azimuth are obtained by constructing an atomic norm minimization problem. In order to solve the atomic norm minimization problem, a two-dimensional mapping operator is proposed, so that a multivariate trigonometric polynomial constraint can be constructed, so that the bounded lemma can be applied to convert the ANM problem into an SDP problem. In order to solve the problem of excessive computational complexity of the 2DMFANM method, the present invention further proposes a fast algorithm 2DMFDANM, which accelerates the calculation by decoupling the spatial angular frequency information.

[0109] Principle description of the joint elevation-azimuth estimation method for broadband systems:

[0110] The signal processed by the method is a broadband far-field plane wave signal received by a uniform rectangular array (URA). The function is to obtain the elevation angle and azimuth angle estimation of the signal source. The schematic diagram is shown in Figure 1 shown.

[0111] The first step of the method is to perform a fast Fourier transform (FFT) on the data received by the uniform rectangular array to convert the signal into a multi-frequency form. The second step is to obtain two matrices with special structures containing the pitch angle and azimuth angle information through the decoupled atomic norm minimization algorithm. The third step is to estimate the spatial angular frequencies from the matrices obtained in the previous step through the information extraction algorithm. The fourth step is to match the spatial angular frequencies through the matching algorithm. The fifth step is to convert the matched spatial angular frequencies into pitch angles and azimuth angles. The algorithm flow is as follows Figure 2 shown.

[0112] Example 1

[0113] The broadband elevation angle-azimuth angle joint estimation method based on the atomic norm minimization method specifically includes:

[0114] Consider r far-field broadband plane waves irradiating an N1×N2 uniform rectangular array (URA). The array has N1 and N2 elements in the x-axis and y-axis directions, respectively. The elevation and azimuth angles of the i-th wave are θ i and φ i . For broadband sources, N f The signal can be divided into N fast Fourier transform (FFT). f subbands, each with a central frequency of f·f0, where f∈{1,2,...,N f}, f0 is the sampling frequency / N f The i-th signal s i can be expressed as

[0115]

[0116] in ‖·‖ is the l2 norm. Here c i represents the amplitude of the i-th signal, Contains N f Let J be a set of array element indices, defined as J = {0, 1, ..., N1-1} × {0, 1, ..., N2-1}. The signal of the f-th subband received by the k-th array element can be expressed as:

[0117]

[0118] Where k∈J, c represents the wave propagation velocity, and d represents the array element spacing.

[0119] Define the spatial angular frequency vector of the i-th signal Can get

[0120]

[0121] It is worth noting that the angular information including the pitch angle and the azimuth angle can be estimated by i Get. f The signal of the kth array element can be organized as follows:

[0122]

[0123] where ⊙ denotes the Khatri-Rao product.

[0124] Combining the signals of the fth subband of all array elements together, we can get the following matrix representation:

[0125]

[0126] It can be further organized as:

[0127]

[0128] in represents the steering vector,

[0129] The matrix A(f,ω i ) is defined as: A(f,ω i )=a2(f,ω 2i )·a1(f,ω 1i ) T .

[0130] Vectorizing (5) yields the following representation for the f-th subband:

[0131]

[0132] in represents the Kronecker product.

[0133] Combining (4) and (6), the complete signal model can be obtained as follows:

[0134]

[0135] in

[0136] The present invention is achieved by To obtain the DOA information to be extracted. In the absence of noise, the method directly processes In practical application scenarios with noise, the received signal can be modeled as

[0137]

[0138] Where N represents the noise.

[0139] Based on (7), define the following atomic set

[0140]

[0141] In order to extract DOA information from the observed signal, the following atomic norm minimization problem is constructed:

[0142] For the noise-free case:

[0143]

[0144] For noisy signals:

[0145]

[0146] Here the atomic norm Defined as:

[0147]

[0148] Where t represents the stretching constant; inf{·} represents the infimum; conv(·) represents the convex hull; and β represents the noise power.

[0149] Since the atomic norm cannot be solved directly, it is necessary to convert the optimization problem into a semi-definite programming (SDP) problem that can be planned and solved. For the sake of brevity, the following will only give the SDP conversion process of the noise-free case (10) in detail, and the noisy case can be obtained in the same way.

[0150] The dual problem of (10) is

[0151]

[0152] in represents the dual norm of the atomic norm in (12), defined as

[0153]

[0154] Where sup means taking the supremum; using (7) Definition and definition We can get:

[0155]

[0156] make We can get:

[0157]

[0158] In order to express (16) in the form of multivariate trigonometric polynomials to use the bounded lemma, the following mapping operator is introduced. Let:

[0159]

[0160] according to The definitions are:

[0161]

[0162] In order to construct the mapping operator, the following definition is made:

[0163]

[0164] in

[0165] make:

[0166]

[0167] as well as

[0168]

[0169] in

[0170] According to (18)-(21), we can see Can be seen as Therefore, we define the mapping operator as follows:

[0171]

[0172] make:

[0173]

[0174] From (7), (22) and (23) we can see that make:

[0175]

[0176] in for Substituting (24) into the definition of Ψ(Q,ω) yields:

[0177] Ψ(Q,ω)=H H z (25)

[0178] Therefore (16) can be transformed into:

[0179]

[0180] Inferred:

[0181]

[0182] Note that the right side of (27) is a multivariate trigonometric polynomial inequality. According to the bounded lemma and the appropriate sum-of-squares relaxation (SOS), this inequality can be written in the following SDP form:

[0183]

[0184] in is a half space, is a The Toeplitz matrix of The diagonal is 1, the rest are 0, i=1,2, N f ×N f The identity matrix of .

[0185] Therefore (13) can be transformed into the following SDP problem:

[0186]

[0187] in represents the conjugate transpose.

[0188] The SDP form of the original problem (10) can be obtained by calculating the dual problem of (29), and the result is as follows:

[0189]

[0190] Among them Toep 2D (μ) is a second-order multi-level Toeplitz matrix,

[0191] The spatial angular frequency vector ω=[ω1,ω2] can be obtained by Toep 2D (μ) is obtained by the MaPP method. Finally, the elevation angle and azimuth angle can be obtained by the following formulas:

[0192]

[0193] It is worth mentioning that the current common SDP solvers, such as SDPT3, can give the solution to the dual problem at the same time after solving the original problem. In practice, since the solution of the ANM dual problem (29) is faster, it is recommended to directly solve (29). Toep in (30) 2D (μ) is the dual variable of P0 in (29) and can be directly given by the solver. Here the SDP form of the dual problem in the presence of noise is given as follows:

[0194]

[0195] Where β is the noise power; ‖·‖2 represents the 2-norm of the matrix.

[0196] The 2DMFANM method flow is shown in Table 1.

[0197] Table 12DMFANM method

[0198]

[0199]

[0200] The improvement of 2DMFDANM method over 2DMFANM method is as follows: In order to derive the decoupling algorithm, the signal is first rewritten as follows:

[0201]

[0202] definition as follows:

[0203]

[0204] Constructing a mapping operator So that:

[0205]

[0206] structure So that:

[0207]

[0208] So we have:

[0209]

[0210] Based on the above, by decoupling (30), we get the following optimization problem:

[0211]

[0212] Where Y = [Y 1 ,...,Y Nf ], T2 is The Toeplitz matrix, YesN f ×N f A block matrix where each block is The spatial angular frequencies ω1 and ω2 can be obtained from T2 and Specifically, ω2 can be obtained by performing Vandermonde decomposition on T2. Has the following structure:

[0213]

[0214] definition:

[0215]

[0216] have:

[0217]

[0218] in where e jis a vector whose elements are all 0 except the jth element which is 1. The obtained S(T1) is a block Toeplitz matrix. By performing Vandermonde decomposition of the block Toeplitz matrix, a decomposition of the form (41) can be obtained, from which ω1 can be extracted. When noise exists, the corresponding optimization problem is as follows:

[0219]

[0220] Where ∈ is the noise power.

[0221] Since {ω1} and {ω2} obtained by the decoupling algorithm are independent, a matching algorithm needs to be designed to match them. Assume that there is an estimate obtained by decoupling and Where K1 and K2 are the numbers of estimated spatial angular frequencies. According to (5), we have:

[0222]

[0223] in, Further: Notice The non-zero elements of have the same index, so the following matching algorithm is designed:

[0224]

[0225] in‖·‖ 2,1 for l 2,1 norm, defined as To optimize the results, Normalize each column to get make By extracting The index corresponding to the largest r elements can achieve matching. The algorithm flow chart of the matching algorithm is shown in Table 2.

[0226] Table 2 Matching algorithm

[0227]

[0228]

[0229] The 2DMFDANM method flow is shown in Table 3.

[0230] Table 32DMFDANM method

[0231]

[0232]

[0233] To illustrate the effectiveness of the method of the present invention, we conducted the following simulation experiments. Consider a 5×5 URA, where the azimuth and elevation angles of the signal are set to (-60°, -45°, -30°) and (30°, 40°, 45°), respectively. Set the wave velocity to v = 1500 m / s. Set the fundamental frequency f0 in the multi-frequency signal model to 1000 Hz. Set the array element spacing to half of the maximum wavelength, i.e. Assume that the signals defined in (1) are independent and follow a complex Gaussian distribution. Assume that the noise is also Gaussian white noise. The optimization problem in the simulation is solved using CVX. The following results are obtained through 100 Monte Carlo experiments.

[0234] In this experiment, we mainly compare the 2DMFANM method with the sparse Bayesian (SBL) algorithm and the traditional ANM-based narrowband algorithm to show its effectiveness. f =5. When there is no noise, the estimated result is as follows Figure 3(a)-Figure 3(f) As shown. Figure 3(d), Figure 3(e) and Figure 3(f) compare the estimation performance of the 2DMFANM method, the SBL method and the traditional method for the spatial angular frequency, while Figure 3(a), Figure 3(b) and Figure 3(c) compare the estimation performance of the pitch angle and azimuth angle. Figure 3(a)-Figure 3(f) It can be seen that the 2DMFANM method uses multi-frequency information to concentrate the estimated results near the true value, while the estimated results of the traditional method are relatively scattered, showing the superiority of the 2DMFANM method.

[0235] To demonstrate the performance of the 2DMFANM method in the presence of noise, the signal-to-noise ratio (SNR) was set to -2dB to 20dB for simulation. When the root mean square error (RMSE) of the spatial angular frequency estimation is less than 0.2, the estimation is considered successful. Figure 4(c) shows the number of successful estimations of the three methods. When calculating the estimated RMSE, only successful estimation results are considered. Figures 4(a) and 4(b) show the curves of the estimated RMSE as a function of SNR. As can be seen from Figure 4(c), the number of successes of all methods increases with the increase of SNR. Compared with the narrowband algorithm (ANM), the 2DMFANM method has a higher success rate at low SNR, indicating that the performance can be improved by using multi-frequency information. As can be seen from Figure 4(a), the performance of the 2DMFANM method for spatial angular frequency estimation at low SNR is not significantly better than that of the SBL method. However, as can be seen from Figure 4(b), the 2DMFANM method significantly outperforms the SBL method in the estimation performance of the elevation angle and azimuth angle. This is mainly because the correspondence between spatial angular frequency and pitch angle and azimuth angle is nonlinear. Since the SBL method is gridded, the estimation results all fall on a fixed grid. The proposed gridless method makes the estimation results continuously gather near the true value. Therefore, when the spatial angular frequency is converted into pitch angle and azimuth angle, the 2DMFANM method will achieve better estimation results. The advantage of gridless can also be reflected in the RMSE of the noise-free case. At this time, the RMSE of the 2DMFANM method is much lower than that of SBL.

[0236] In order to verify the performance of the proposed fast algorithm 2DMFDANM, consider the noise-free case and set N f Set it to 1-5, and simulate the original algorithm 2DMFANM and the fast algorithm 2DMFDANM. Figure 5(a)-Figure 5(d) Shows when N f The estimation results of each algorithm when =5.

[0237] When the RMSE of the spatial angular frequency estimation is less than 0.05, the estimation is considered successful. The number of successful estimations is summarized in Table 4. It can be seen that the success rates of the 2DMFANM method and the fast algorithm 2DMFDANM increase with N f significantly increased with the increase of .

[0238] Table 4 Number of successful estimations by fast algorithm

[0239]

[0240] The operation time of the two algorithms is summarized in Table 5. It can be seen that the operation time of the original 2DMFANM algorithm decreases with N f 2DMFDANM has the ability to reduce the operation time.

[0241] Table 5 Fast algorithm operation time unit: second

[0242]

[0243] The RMSE of the two algorithms for pitch angle-azimuth estimation and RMSE for spatial angular frequency estimation are summarized in Table 6 and Table 7 respectively. It can be seen that as N f As N increases, the estimation accuracy of 2DMFANM increases significantly. The estimation accuracy of the fast algorithm 2DMFDANM increases with N f The performance of the fast algorithm can be summarized as follows: when computing resources are insufficient, 2DMFDANM can significantly improve the computing speed at the expense of some estimation accuracy.

[0244] Table 6 RMS of the fast algorithm to estimate the pitch angle and azimuth angle

[0245]

[0246] Table 7 RMSE of fast algorithm for estimating spatial angular frequency

[0247]

[0248] Example 2

[0249] The present application also provides a pitch angle-azimuth angle joint estimation device applicable to a broadband system, which is implemented based on the above method, and the device includes:

[0250] The Fourier transform module is used to perform fast Fourier transform on the data received by the uniform rectangular array and convert the signal into a multi-frequency form;

[0251] An estimated spatial angular frequency module, for obtaining estimated values ​​of two spatial angular frequencies using a decoupled atomic norm minimization method;

[0252] A spatial angular frequency matching module, used for matching the estimated spatial angular frequency through a matching algorithm;

[0253] The module for calculating the pitch angle and azimuth angle is used to convert the matched spatial angular frequency into the pitch angle and azimuth angle.

[0254] Example 3

[0255] The present application may also provide a computer device, comprising: at least one processor, a memory, at least one network interface and a user interface. The various components in the device are coupled together through a bus system. It is understood that the bus system is used to achieve connection and communication between these components. In addition to the data bus, the bus system also includes a power bus, a control bus and a status signal bus.

[0256] The user interface may include a display, a keyboard or a pointing device, such as a mouse, a trackball, a touch pad or a touch screen.

[0257] It is understood that the memory in the embodiments disclosed in the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct memory bus random access memory (DRRAM). The memories described herein are intended to include, but are not limited to, these and any other suitable types of memories.

[0258] In some embodiments, the memory stores the following elements, executable modules or data structures, or a subset thereof, or an extended set thereof: an operating system and applications.

[0259] The operating system includes various system programs, such as a framework layer, a core library layer, a driver layer, etc., which are used to implement various basic services and process hardware-based tasks. The application includes various application programs, such as a media player (Media Player), a browser (Browser), etc., which are used to implement various application services. The program for implementing the method of the embodiment of the present disclosure can be included in the application.

[0260] In the above embodiment, the processor may also call a program or instruction stored in the memory, specifically, a program or instruction stored in an application program, and is used to:

[0261] Execute the steps of the above method.

[0262] The above method can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software. The above processor may be a general processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The above disclosed methods, steps and logic block diagrams can be implemented or executed. The general processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the above disclosed method can be directly embodied as a hardware decoding processor to execute, or the hardware and software modules in the decoding processor can be executed. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0263] It is understood that the embodiments described in the present application can be implemented by hardware, software, firmware, middleware, microcode or a combination thereof. For hardware implementation, the processing unit can be implemented in one or more application specific integrated circuits (ASIC), digital signal processors (DSP), digital signal processing devices (DSPD), programmable logic devices (PLD), field programmable gate arrays (FPGA), general-purpose processors, controllers, microcontrollers, microprocessors, other electronic units for performing the functions described in the present application or a combination thereof.

[0264] For software implementation, the technology of the present application can be implemented by executing the functional modules (such as procedures, functions, etc.) of the present application. The software code can be stored in a memory and executed by a processor. The memory can be implemented in the processor or outside the processor.

[0265] Example 4

[0266] The present application may also provide a non-volatile storage medium for storing a computer program. When the computer program is executed by a processor, each step in the above method embodiment can be implemented.

[0267] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present application and are not intended to limit it. Although the present application is described in detail with reference to the embodiments, a person skilled in the art should understand that any modification or equivalent replacement of the technical solution of the present application does not depart from the spirit and scope of the technical solution of the present application and should be included in the scope of the claims of the present application.

Claims

1. A joint elevation angle-azimuth angle estimation method applicable to a broadband system, comprising: Step 1: Perform fast Fourier transform on the data received by the uniform rectangular array to convert the signal into a multi-frequency form; Step 2: Use the decoupled atomic norm minimization method to obtain estimates of the two spatial angular frequencies; Step 3: Match the estimated spatial angular frequency through a matching algorithm; Step 4: Convert the matched spatial angular frequency into elevation angle and azimuth angle.

2. The method for joint elevation-azimuth estimation suitable for broadband systems according to claim 1, characterized in that: The step 2 comprises: For the noise-free case, the following atomic norm minimization problem is constructed: in, represents the received signal without noise; Y represents the received signal containing noise; Among them, c i represents the amplitude of the i-th signal; r represents the number of far-field broadband plane wave signals; ‖·‖ is the l2 norm, N f Indicates the number of subbands a signal is divided into; ⊙ indicates the Khatri-Rao product; a1(f,ω 1i )=[1,exp(j2πfω 1i ),...,exp(j2πfω 1i ·(N1-1));a2(f,ω 2i )=[1,exp(j2πfω 2i ),...,exp(j2πfω 2i ·(N2-1));ω i represents the spatial angular frequency vector of the i-th signal, ω 1i Represents ω i Horizontal component; ω 2i Represents ω i The vertical component; N1 represents the number of array elements in the horizontal direction; N2 represents the number of array elements in the vertical direction; f0 represents the sampling frequency / N f ; c represents the wave propagation speed, d represents the array element spacing; θ i represents the pitch angle of the i-th signal; φ i represents the azimuth of the i-th signal; T represents vector transposition; Atomic norm Defined as: Where t is the stretching constant; inf{·} is the infimum; conv(·) is the convex hull; the atomic set Defined as: Using the decoupled atomic norm minimization method, the following optimization problem is obtained: in, T2 is Toeplitz matrix of ; Tr(·) means taking the trace of the matrix; represents conjugate transpose; definition as follows: Mapping Operator is defined as follows: So that: N f ×N f A block matrix where each block is The Toeplitz matrix: K1 is the number of horizontal components of the estimated spatial angular frequency; definition Depend on Get S(T1), extract ω from S(T1) 1i ; in, where e j is a vector whose elements are all 0 except the jth element which is 1; By performing Vandermonde decomposition on T2, we can obtain ω 2i .

3. The method for joint elevation-azimuth estimation suitable for broadband systems according to claim 1, characterized in that: The step 2 comprises: For the noisy case, the following atomic norm minimization problem is constructed: in, represents the received noise-free signal; Y represents the received signal containing noise; β represents the noise power; ‖·‖2 represents the 2-norm of the matrix; Among them, c i represents the amplitude of the i-th signal; r represents the number of far-field broadband plane wave signals; ‖·‖ is the l2 norm, N f Indicates the number of subbands a signal is divided into; ⊙ indicates the Khatri-Rao product; a1(f,ω 1i )=[1,exp(j2πfω 1i ),...,exp(j2πfω 1i ·(N1-1));a2(f,ω 2i )=[1,exp(j2πfω 2i ),...,exp(j2πfω 2i ·(N2-1));ω i represents the spatial angular frequency vector of the i-th signal, ω 1i Represents ω i Horizontal component; ω 2i Represents ω i The vertical component; N1 represents the number of array elements in the horizontal direction; N2 represents the number of array elements in the vertical direction; f0 represents the sampling frequency / N f ; c represents the wave propagation speed, d represents the array element spacing; θ i represents the pitch angle of the i-th signal; φ i represents the azimuth of the i-th signal; T represents vector transposition; Atomic norm Defined as: Where t is the stretching constant; inf{·} is the infimum; conv(·) is the convex hull; the atomic set Defined as: Using the decoupled atomic norm minimization method, the following optimization problem is obtained: Where, ∈ is the noise power; Y = [Y 1 ,...,Y Nf ], T2 is Toeplitz matrix of ; Tr(·) means taking the trace of the matrix; represents conjugate transpose; definition as follows: Mapping Operator is defined as follows: So that: N f ×N f A block matrix where each block is The Toeplitz matrix: K1 is the number of horizontal components of the estimated spatial angular frequency; definition Depend on Get S(T1), extract ω from S(T1) 1i ; in, where e j is a vector whose elements are all 0 except the jth element which is 1; By performing Vandermonde decomposition on T2, we can obtain ω 2i .

4. The method for joint elevation-azimuth estimation applicable to broadband systems according to claim 2 or 3, characterized in that: The step 3 comprises: The matching algorithm is: f=1,...N f in,‖·‖ 2,1 for l 2,1 norm; vec(·) means converting the matrix into a vector; K2 is the number of vertical components of the estimated spatial angular frequency; and are the estimated values ​​of the horizontal and vertical components of the spatial angular frequency, respectively; To optimize the results, Normalize each column to get set up By extracting The indices corresponding to the largest r elements can be matched.

5. The method for joint elevation-azimuth estimation suitable for broadband systems according to claim 4, characterized in that: The step 4 comprises: The pitch angle θ in the i-th signal i and azimuth φ i They are obtained by the following formulas:

6. A pitch angle-azimuth angle joint estimation device suitable for a broadband system, implemented based on the method described in any one of claims 1 to 5, characterized in that: The device comprises: The Fourier transform module is used to perform fast Fourier transform on the data received by the uniform rectangular array and convert the signal into a multi-frequency form; An estimated spatial angular frequency module, for obtaining estimated values ​​of two spatial angular frequencies using a decoupled atomic norm minimization method; A spatial angular frequency matching module, used to match the estimated spatial angular frequency through a matching algorithm; and The module for calculating the pitch angle and azimuth angle is used to convert the matched spatial angular frequency into the pitch angle and azimuth angle.

7. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 5 is implemented.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, causes the processor to perform the method according to any one of claims 1 to 5.

Citation Information

Patent Citations

  • Atomic norm minimization-based dimensionality-reducible two-dimensional out-of-lattice DOA estimation method

    CN113673317A

  • Fast meshless direction-of-arrival estimation method suitable for broadband linear frequency modulation signal

    CN117932319A

  • Meshless DOA estimation method based on L-type co-prime array

    CN117949889A

  • Method, system, and intelligent terminal for one-bit quantization direction of arrival estimation

    US12044786B1

Cited By

  • Mesh-free DOD-DOA joint estimation method and system based on Smoothed ANM

    CN120161405A

  • A Meshless DOD-DOA Joint Estimation Method and System Based on Smoothed ANM

    CN120161405B