A shallow sea array grouping weighted DOA estimation method based on horizontal line array

Abstract

This invention belongs to the field of underwater acoustic signal processing and analysis, and discloses a weighted DOA estimation method for shallow-sea arrays based on horizontal linear arrays. Specifically, a narrowband pulse sound source is emitted in a shallow-sea waveguide. After passing through the shallow-sea acoustic channel, the signal is received at the receiving point. The linear array is split according to different element spacings, dividing one array into many groups with different element spacings. DOA estimation is performed using the data received from each group of arrays. Finally, the estimation results of each group of arrays are weighted using a Taylor series method to obtain the final DOA estimate. After eigenvalue decomposition, the noise subspace is corrected. This method, given the limited number of arrays, makes full use of the array data, significantly improving the accuracy of DOA estimation.

Description

Technical Field

[0001] This invention belongs to the field of underwater acoustic signal processing and analysis, specifically relating to a DOA estimation method based on array grouping weighted subspace. Background Technology

[0002] Direction of arrival (DOA) estimation is one of the most critical parameters in target parameter estimation, playing a vital role in underwater target localization, underwater communication, marine resource surveying, and marine archaeology. With the development of underwater acoustic communication technology, DOA estimation of underwater acoustic array signals has become a prerequisite for research on underwater acoustic target identification, localization, tracking, and detection. However, compared to spatial electromagnetic channels, underwater acoustic channels are far more complex, and traditional spatial DOA algorithms cannot guarantee good results in underwater acoustic array DOA estimation, exhibiting significant limitations. Summary of the Invention

[0003] To address the shortcomings of existing DOA algorithms in underwater acoustic array DOA estimation, such as poor accuracy and slow speed, this invention provides a high-precision DOA estimation method for narrowband pulses in shallow sea waveguides that maximizes array data availability, is easy to operate, and has a fast computation speed.

[0004] To achieve the above objectives, the present invention is implemented through the following technical solution:

[0005] This invention is a grouped weighted DOA estimation method for shallow sea arrays based on horizontal linear arrays, comprising the following steps:

[0006] Step a: A narrowband pulse sound source S is emitted in a shallow sea waveguide. The seabed conditions of the shallow sea waveguide are a hard and uniform seabed. After the signal source passes through the shallow sea sound channel, the received signal X is obtained at the receiving point.

[0007] Step b: The received signal X is grouped according to the array element interval to obtain the received data of each array group as X1, X2, ..., Xk (k is the sequence number, k = 1, 2, 3, ...);

[0008] Step c: Calculate the covariance matrix of each group according to formula (1), and then perform eigenvalue decomposition on the covariance matrix to obtain the signal subspace Us and the noise subspace Un;

[0009]

[0010] In equation (1), R k Let be the covariance matrix, (.) H Let E(.) denote the conjugate transpose of a matrix, and E(.) denote the expectation. s =diag(λ1,λ2,...,λ) i ) is a signal feature matrix composed of the i largest eigenvalues, ΛN =diag(λ i+1 ,λ i+2 ,...,λ M ) is a noise feature matrix composed of smaller eigenvalues, where λ represents the eigenvalue. U s =[u1,u2,...,u i ], U N =[u i+1 ,u i+2 ,...,u M ], u represents the eigenvector. i is the number of signal sources, M is the number of array elements, and diag(.) means generating a diagonal matrix using the values ​​in parentheses.

[0011] Step d: Correct the obtained noise subspace Un according to equation (2) to obtain the constrained weighted subspace C. n ;

[0012]

[0013] Step e: Following the MUSIC algorithm, solve for the spatial spectrum function P of each group of signals using the corrected noise subspace. k-MUSIC ;

[0014]

[0015] In the formula, A(θ) is the array steering matrix.

[0016] Step f involves weighting the spatial spectrum functions obtained from each group using Taylor series to obtain the final signal spatial spectrum function P. MUSCI A peak search is performed on the spatial spectral function, and the direction corresponding to the peak is the estimated value of DOA.

[0017]

[0018] In equation (4), P MUSCI For the final spatial spectral function, P k-MUSIC Let be the spatial spectral functions for each group, k = 1, 2, ..., n, and (.)! denotes factorial.

[0019] Furthermore, in step a, the narrowband pulse sound source; the sound source propagation environment is a shallow sea waveguide, and its specific environmental conditions are: the waveguide characteristics do not change in the horizontal direction, the waveguide is horizontally layered, and it has a hard, uniform, high-sound-velocity seabed; the sound source location is any location in the seawater; the receiver is a horizontal linear array receiver; the receiver location is any location in the seawater.

[0020] Furthermore, in step b, array elements are selected in a manner that the spacing between adjacent array elements continuously increases.

[0021] The beneficial effects of this invention are:

[0022] This invention can make full use of existing arrays to receive data and improve the accuracy of DOA estimation;

[0023] This invention does not require an increase in the number of array elements, thus reducing hardware costs;

[0024] This invention is easy to operate, has a simple algorithm, requires little data processing, has a fast calculation speed, and has high accuracy. Attached Figure Description

[0025] Figure 1 This is a flowchart of the estimation method of the present invention.

[0026] Figure 2 This is a comparison diagram of the azimuth spatial spectrum of the method of the present invention and the traditional method. Detailed Implementation

[0027] The embodiments of the present invention will be disclosed below with reference to the drawings. For clarity, many practical details will be described in the following description. However, it should be understood that these practical details are not intended to limit the invention. That is, in some embodiments of the invention, these practical details are not essential.

[0028] like Figure 1 As shown, this invention is a grouped weighted DOA estimation method for shallow sea arrays based on horizontal linear arrays. The method includes the following steps:

[0029] Step 1: A narrowband pulse sound source S is emitted in a shallow sea waveguide. The signal source of the narrowband pulse sound source S is received at the receiving point after passing through the shallow sea sound channel.

[0030] The sound source propagation environment of the narrowband pulse sound source S is a shallow sea waveguide. Specifically, the waveguide characteristics do not change in the horizontal direction, the waveguide is horizontally layered, and it has a hard, uniform, high-sound-velocity seabed. The sound source location is any location in the seawater, and the receiver is a horizontal linear array receiver located at any location in the seawater.

[0031] Step 2: The received signal X from Step 1 is grouped into arrays according to the array element interval. The array elements are selected in such a way that the interval between adjacent array elements increases continuously. The received data of each array group is represented as X1, X2, ..., Xk, where k is the sequence number and k≥1.

[0032] Step 3: Calculate the covariance matrix of each array group in Step 2 using formula (1), and perform eigenvalue decomposition on the covariance matrix to obtain the signal subspace Us and the noise subspace Un.

[0033]

[0034] In the formula R k Let be the covariance matrix, (.) H E(.) denotes the conjugate transpose of a matrix, and E(.) denotes the expectation. s =diag(λ1,λ2,...,λ) i ) is a signal feature matrix composed of the i largest eigenvalues, Λ N =diag(λ i+1 ,λ i+2 ,...,λ M ) is a noise feature matrix composed of smaller eigenvalues, where λ represents the eigenvalue, U s =[u1,u2,...,u i ], U N =[u i+1 ,u i+2 ,...,u M ], u represents the eigenvector, i is the number of signal sources, M is the number of array elements, and diag(.) means generating a diagonal matrix using the values ​​in parentheses.

[0035] Step 4: Use formula (2) to correct the noise subspace Un obtained in step 3 to obtain the constrained weighted subspace C. n .

[0036]

[0037] Step 5: Using the MUSIC algorithm, solve for the spatial spectrum function P of each group of signals using the noise subspace Un corrected in Step 4. k-MUSIC ;

[0038]

[0039] In the formula, A(θ) is the array steering matrix, and C n To constrain the weighted noise subspace, θ is an angle;

[0040] Step 6: Weight the spatial spectrum functions obtained from each group using Taylor series to obtain the final signal spatial spectrum function P. MUSCI ,

[0041]

[0042] In the formula P MUSCI For the final spatial spectral function, P k-MUSIC Let k be the spatial spectral function for each group, k = 1, 2, ..., n, and (.)! denote factorial;

[0043] A peak search is performed on the spatial spectral function, and the direction corresponding to the peak is the DOA estimate.

[0044] Figure 2The figure shows a comparison of the DOA estimation results of the estimation method of this invention and the prior art, with the spectral peak direction representing the DOA estimate. Simulation comparisons were performed for different numbers of information sources, different signal-to-noise ratios, and different numbers of snapshots.

[0045] Figure 2 (a) and (b) are simulations under different signal-to-noise ratios. Except for the signal-to-noise ratio, all other conditions are the same. It can be seen that the resolution and accuracy of the MUSIC algorithm of this invention are higher than those of the existing MUSIC algorithms.

[0046] Figure 2 (a), (c), and (d) represent simulations with different numbers of information sources, where all other conditions are the same except for the number of information sources. From single-source to multi-source simulations, the estimation method proposed in this invention demonstrates better performance. Figure 2 (d)(c)(f) represent simulations under different number of snapshots. Except for the number of snapshots, all other conditions are the same. The method of the present invention has better performance under different number of snapshots.

[0047] In summary, the results demonstrate that the weighted DOA estimation method for shallow sea arrays based on horizontal linear arrays in this invention has higher accuracy.

[0048] The above description is only a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. Any equivalent modifications or changes made by those skilled in the art based on the content disclosed in the present invention should be included within the scope of protection set forth in the claims.

Claims

1. A shallow water array grouping weighted DOA estimation method based on horizontal line array, characterized in that: The shallow sea array grouping weighted DOA estimation method comprises the following steps: Step 1, a narrow-band pulse sound source S is transmitted in a shallow sea waveguide, the seabed condition of the shallow sea waveguide is hard and uniform, and a signal source of the narrow-band pulse sound source S is received as a receiving signal X after passing through a shallow sea sound channel; Step 2, the receiving signal X of step 1 is array grouped according to an array element interval, and receiving data of each group array is represented as X1, X2, …, Xk, wherein k is a serial number, and k is greater than or equal to 1; Step 3, a covariance matrix of each group array in step 2 is obtained, eigenvalue decomposition is performed on the covariance matrix, a signal subspace Us and a noise subspace Un are obtained, Step 4, revising the noise subspace Un obtained in step 3 to obtain a constraint weighted subspace C n ; Step 5, according to the MUSIC algorithm, the spatial spectrum function of each group signal is solved by using the noise subspace Un corrected in step 4; Step 6, the spatial spectrum function solved by each group is weighted and processed according to a Taylor series, and a final signal spatial spectrum function is obtained; Step 7, a spectrum peak search is performed on the spatial spectrum function, and a direction corresponding to the spectrum peak is a DOA estimation value, wherein The spatial spectrum function P of each group signal in step 5 k-MUSIC is where A(θ) is the array steering matrix, C n is the constrained weighted noise subspace, and θ is the angle. The step 6 gives the final signal space spectrum function P MUSCI is where P MUSCI is the final spatial spectrum function, P k-MUSIC is the group spatial spectrum function, k = 1, 2,..., n, and (.)! denotes factorial.

2. The shallow water array grouping weighted DOA estimation method based on horizontal line array according to claim 1, characterized in that: In the step 1, the sound source propagation environment of the narrow-band pulse sound source S is a shallow sea waveguide.

3. The shallow water array grouping weighted DOA estimation method based on horizontal line array according to claim 2, characterized in that: The sound source propagation environment of the narrow-band pulse sound source S is specifically that a waveguide feature does not change in a horizontal direction, the waveguide is horizontally layered, has a hard and uniform high sound speed seabed, a sound source position is any position in seawater, receiving is horizontal line array receiving, and a receiving position is any position in seawater.

4. The shallow water array grouping weighted DOA estimation method based on horizontal line array according to claim 3, characterized in that: In the step 2, the array grouping is selected according to an increasing manner of adjacent element intervals.

Images