DOA estimation method and device based on non-uniform linear array, computer equipment, medium and product

Abstract

The invention relates to a DOA estimation method and device based on a non-uniform linear array, computer equipment, a medium and a product. The method comprises the following steps: establishing a signal model of a source signal and a non-uniform linear array, dividing the non-uniform linear array into a plurality of sub-arrays according to an interval relationship between array elements, sequentially collecting sub-array signals of a target sub-array corresponding to a sampling sequence according to a preset sampling sequence, establishing a signal covariance matrix according to the signal model and all the sub-array signals, and obtaining a signal covariance matrix of the target sub-array according to the signal covariance matrix. Performing approximate processing on each element in the signal covariance matrix to obtain an approximate covariance matrix, determining index positions of the approximate covariance matrix, solving the approximate covariance matrix according to all the index positions to obtain a target covariance matrix, performing characteristic decomposition on the target covariance matrix to obtain a target covariance matrix; and DOA estimation is carried out based on a characteristic decomposition result. By adopting the method, the direction information of the signal can be more accurately extracted from the noise background, and the DOA estimation accuracy is improved.

Description

Technical Field

[0001] The present application relates to the technical field of array sensors, and in particular to a DOA estimation method, device, computer equipment, medium and product based on a non-uniform linear array. Background Art

[0002] Direction of Arrival (DOA) estimation is an important research topic in array signal processing, which aims to determine the direction of the source signal relative to the receiving array. For example, in radar, sonar and other systems, by receiving the signal transmitted by the target, the DOA estimation technology is used to determine the azimuth, pitch angle and other direction information of the target relative to the receiving antenna array or sensor array.

[0003] At present, in the construction of theoretical models for DOA estimation, it is usually assumed that the array elements are independent of each other. However, in an antenna array, when the array element spacing is small, the electromagnetic fields between the array elements will affect each other, resulting in mutual coupling. The mutual coupling effect will cause the actual array response to be different from the array response under ideal conditions, which will lead to a mismatch between the model and the actual situation, affecting the accuracy of DOA estimation. Summary of the invention

[0004] Based on this, it is necessary to provide a DOA estimation method, device, computer equipment, medium and product based on a non-uniform linear array that can accurately obtain DOA estimation in response to the above technical problems.

[0005] In a first aspect, the present application provides a DOA estimation method based on a non-uniform linear array, comprising:

[0006] Establish the signal model of source signal and non-uniform linear array;

[0007] According to the spacing relationship between array elements, the non-uniform linear array is divided into multiple sub-arrays, and the sub-array signals of the target sub-array corresponding to the sampling order are collected in sequence according to the preset sampling order; the target sub-array corresponding to each sampling order is in a normal working state, and the sub-arrays other than the target sub-array are in a non-working state;

[0008] A signal covariance matrix is ​​established according to the signal model and all sub-array signals, and each element in the signal covariance matrix is ​​approximated to obtain an approximate covariance matrix;

[0009] Determine the index position of the approximate covariance matrix, solve the approximate covariance matrix according to all index positions, and obtain the target covariance matrix;

[0010] The target covariance matrix is ​​subjected to eigendecomposition, and DOA estimation is performed based on the result of eigendecomposition.

[0011] In one embodiment, the step of establishing a signal model of a source signal and a non-uniform linear array includes:

[0012] According to the array element position and array operating wavelength of the non-uniform linear array, the DOA expression corresponding to each source signal is determined, and the array signal expression is determined according to the DOA expressions corresponding to all source signals. The signal expression is used to characterize the ideal array signal received in the absence of mutual coupling effect.

[0013] In the case of the existence of mutual coupling effect, a mutual coupling matrix is ​​established, and a signal model is established according to the mutual coupling matrix and the array signal expression, and the signal model represents the array signal received in the case of the existence of the mutual coupling effect.

[0014] In one embodiment, the step of dividing the non-uniform linear array into a plurality of sub-arrays according to the spacing relationship between array elements includes:

[0015] Determine the mutual coupling range according to the operating wavelength of the non-uniform linear array;

[0016] The non-uniform linear array is divided into a plurality of sub-arrays according to the mutual coupling range, and there is no mutual coupling effect between the array elements in each sub-array.

[0017] In one embodiment, the step of performing approximate processing on each element in the signal covariance matrix to obtain an approximate covariance matrix includes:

[0018] For each element in the signal covariance matrix, the element is decomposed by the Bessel function to obtain the Bessel combination expression corresponding to the element;

[0019] The number of approximation terms corresponding to the element is determined, and the Bessel combination expression is converted into an approximate expression according to the number of approximation terms to obtain the corresponding approximate covariance matrix.

[0020] In one embodiment, the step of solving the approximate covariance matrix according to all index positions to obtain the target covariance matrix includes:

[0021] According to the matrix structure of the approximate covariance matrix, an objective function is established, and the objective function includes the trace of the matrix, the regularization parameter and the norm of the matrix;

[0022] Determine constraints based on index position;

[0023] Under the constraints, the objective function is solved to obtain the target covariance matrix.

[0024] In one embodiment, the step of performing eigendecomposition on the target covariance matrix and estimating the DOA based on the result of the eigendecomposition includes:

[0025] Perform eigendecomposition on the target covariance matrix to obtain signal subspace features and noise subspace features;

[0026] Based on the signal subspace characteristics and noise subspace characteristics, a spatial spectrum function is established;

[0027] The DOA of the source signal is determined based on the peak value of the spatial spectrum function.

[0028] In a second aspect, the present application further provides a DOA estimation device based on a non-uniform linear array, comprising:

[0029] A model building module, used to build a signal model of a source signal and a non-uniform linear array;

[0030] The linear array division module is used to divide the non-uniform linear array into multiple sub-arrays according to the spacing relationship between array elements, and collect sub-array signals of the target sub-array corresponding to the sampling sequence in sequence according to the preset sampling sequence; the target sub-array corresponding to each sampling sequence is in a normal working state, and the sub-arrays other than the target sub-array are in a non-working state;

[0031] A matrix calculation module is used to establish a signal covariance matrix according to the signal model and all sub-array signals, and to perform approximate processing on each element in the signal covariance matrix to obtain an approximate covariance matrix;

[0032] A matrix solving module is used to determine the index position of the approximate covariance matrix, solve the approximate covariance matrix according to all the index positions, and obtain the target covariance matrix;

[0033] The DOA estimation module is used to perform eigendecomposition on the target covariance matrix and perform DOA estimation based on the result of the eigendecomposition.

[0034] In a third aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method steps of any one of the first aspects are implemented.

[0035] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, which implements any one of the method steps in the first aspect when the computer program is executed by a processor.

[0036] In a fifth aspect, the present application further provides a computer program product, comprising a computer program, which implements any one of the method steps in the first aspect when executed by a processor.

[0037] The above-mentioned DOA estimation method, device, computer equipment, medium and product based on non-uniform linear array, by establishing a signal model of the source signal and the non-uniform linear array, divides the non-uniform linear array into multiple sub-arrays according to the spacing relationship between the array elements, establishes a signal covariance matrix according to the signal model and all sub-array signals, and performs approximate processing on each element in the signal covariance matrix to obtain an approximate covariance matrix, which can suppress the influence of noise on signal processing to a certain extent, and when finally performing DOA estimation, it can more accurately extract the direction information of the signal from the noise background, thereby improving the accuracy of DOA estimation. In addition, by sequentially collecting the sub-array signals of the target sub-array corresponding to the sampling order according to the preset sampling order, only part of the sub-array works each time, which can reduce the number of hardware devices working at the same time, thereby reducing the hardware complexity. At the same time, through simple space-time mutually exclusive sampling, the mutual coupling effect can be suppressed, thereby effectively reducing the complexity of the calculation process and improving the response speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0039] FIG1 is a diagram showing an application environment of a DOA estimation method based on a non-uniform linear array in one embodiment;

[0040] FIG2 is a schematic flow chart of a DOA estimation method based on a non-uniform linear array in one embodiment;

[0041] FIG3 is a schematic diagram of a source signal incident on a non-uniform linear array in one embodiment;

[0042] FIG4 is a schematic diagram of a sub-array structure divided in one embodiment;

[0043] FIG5 is a scatter plot of a Monte Carlo simulation in one embodiment;

[0044] FIG6 is a schematic flow chart of a DOA estimation method based on a non-uniform linear array in another embodiment;

[0045] FIG7 is a structural block diagram of a DOA estimation device based on a non-uniform linear array in one embodiment;

[0046] FIG. 8 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0048] The DOA estimation method based on non-uniform linear array provided in the embodiment of the present application can be applied to the application environment shown in FIG. 1 . The terminal 102 communicates with the server 104 through a network. The data storage system can store data that the server 104 needs to process. The data storage system can be integrated on the server 104, or placed on a cloud or other network server. The terminal 102 is used to establish a signal model of the source signal and the non-uniform linear array, divide the non-uniform linear array into a plurality of sub-arrays according to the spacing relationship between the array elements, and sequentially collect the sub-array signals of the target sub-array corresponding to the sampling order according to the preset sampling order, establish a signal covariance matrix according to the signal model and all sub-array signals, perform approximate processing on each element in the signal covariance matrix to obtain an approximate covariance matrix, determine the index position of the approximate covariance matrix, solve the approximate covariance matrix according to all the index positions, obtain the target covariance matrix, perform eigendecomposition on the target covariance matrix, and perform DOA estimation based on the result of the eigendecomposition. The terminal 102 may be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart car devices, projection devices, etc. Portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. Head-mounted devices may be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc. The server 104 may be an independent physical server, a server cluster or distributed system consisting of multiple physical servers, or a cloud server that provides cloud computing services.

[0049] In an exemplary embodiment, as shown in FIG2 , a DOA estimation method based on a non-uniform linear array is provided, and the method is applied to the terminal 102 in FIG1 as an example for description, including the following steps 202 to 210. Among them:

[0050] S202: Establish a signal model of the source signal and the non-uniform linear array.

[0051] Optionally, when establishing a signal model of a source signal and a non-uniform linear array, the geometric relationship of signal propagation is involved. For source signals arriving from different directions, the time delay or phase difference reaching each array element of the array is related to the DOA of the signal. The signal model can express this time delay or phase difference in the form of a mathematical expression. For example, under the far-field plane wave assumption, the phase difference between array elements is related to the arrival angle of the signal. This angle-related information is included in the signal model and is the basis for DOA estimation.

[0052] Optionally, the signal model of the non-uniform linear array includes an array response vector, which describes the array's response to incident signals from different directions. When performing DOA estimation, it is necessary to analyze the signal received by the array and find the direction that matches the array response vector to determine the direction of arrival of the signal. For example, in some subspace-based DOA estimation methods (such as the MUSIC algorithm), it is necessary to first construct an array manifold matrix based on the signal model. This matrix is ​​composed of array response vectors corresponding to different DOAs. The signal subspace and noise subspace are separated by eigendecomposing the covariance matrix of the array received signal, and then the relationship between the array manifold matrix and these subspaces is used to estimate the DOA of the signal.

[0053] S204: Divide the non-uniform linear array into a plurality of sub-arrays according to the spacing relationship between array elements, and collect sub-array signals of target sub-arrays corresponding to the sampling order in sequence according to a preset sampling order; the target sub-array corresponding to each sampling order is in a normal working state, and the sub-arrays other than the target sub-array are in a non-working state.

[0054] Optionally, in DOA estimation, for a non-uniform linear array, when the source signal reaches each array element from different directions, due to the non-uniform array element spacing, the time delay and phase difference of the signal reaching each array element will be different due to the DOA. By dividing the non-uniform linear array into multiple sub-arrays, different sub-arrays have different response characteristics to signals in different directional ranges. Afterwards, the sub-array signals are collected in a preset sampling order. At any sampling moment, only the target sub-array is in a normal working state and collects signals, and other sub-arrays are in a non-working state. This can reduce the number of array elements working at the same time, reduce hardware complexity and energy consumption, and is also conducive to phased and targeted processing of signals.

[0055] S206: Establish a signal covariance matrix according to the signal model and all sub-array signals, and perform approximate processing on each element in the signal covariance matrix to obtain an approximate covariance matrix.

[0056] Optionally, the signal covariance matrix is ​​a matrix that describes the correlation between signals. This matrix is ​​constructed based on the previously established signal model and all the collected subarray signals. It contains the relevant information of the signals between different array elements (or subarrays). In actual situations, due to the presence of noise and limited sampling data, the amount of calculation required to accurately calculate the covariance matrix is ​​large. By approximating the covariance matrix, the main information related to DOA can be extracted more effectively.

[0057] S208: Determine the index position of the approximate covariance matrix, and solve the approximate covariance matrix according to all the index positions to obtain a target covariance matrix.

[0058] Optionally, the index positions in the approximate covariance matrix correspond to specific subarray relationships or signal characteristics related to the direction of arrival of the signal. For example, some index positions may be related to the signal correlation between subarrays corresponding to a specific angle range. By determining these index positions, the information most relevant to DOA estimation can be screened out. The approximate covariance matrix is ​​solved according to all the determined index positions to obtain a target covariance matrix. In DOA estimation, the characteristic structure of the target covariance matrix is ​​more directly related to the direction of arrival of the signal.

[0059] S210: Perform eigendecomposition on the target covariance matrix, and perform DOA estimation based on the result of the eigendecomposition.

[0060] Optionally, after eigendecomposing the target covariance matrix, the eigenvalues ​​and eigenvectors obtained are directly related to the DOA. In the subspace-based DOA estimation method, the eigenvectors corresponding to large eigenvalues ​​are related to the signal subspace, and the signal subspace contains information about the direction of arrival of the signal. For example, in the MUSIC algorithm, by constructing a spatial spectrum function, the relationship between the eigenvector and the array manifold is used to search for the direction of arrival of the signal, thereby achieving accurate DOA estimation.

[0061] In the above-mentioned DOA estimation method based on non-uniform linear array, by establishing a signal model of the source signal and the non-uniform linear array, the non-uniform linear array is divided into multiple sub-arrays according to the spacing relationship between the array elements, a signal covariance matrix is ​​established according to the signal model and all sub-array signals, and each element in the signal covariance matrix is ​​approximated to obtain an approximate covariance matrix, which can suppress the influence of noise on signal processing to a certain extent. When the DOA estimation is finally performed, the direction information of the signal can be more accurately extracted from the noise background, thereby improving the accuracy of DOA estimation. In addition, by sequentially collecting the sub-array signals of the target sub-array corresponding to the sampling order according to the preset sampling order, only part of the sub-array works each time, which can reduce the number of hardware devices working at the same time, thereby reducing the hardware complexity. At the same time, through simple space-time mutually exclusive sampling, the mutual coupling effect can be suppressed, thereby effectively reducing the complexity of the operation process and improving the response speed.

[0062] In an exemplary embodiment, the step of establishing a signal model of a source signal and a non-uniform linear array includes: determining a DOA expression corresponding to each source signal according to an array element position and an array operating wavelength of the non-uniform linear array, and determining an array signal expression according to the DOA expressions corresponding to all source signals, the signal expression being used to characterize an ideal array signal received in the absence of a mutual coupling effect; in the presence of a mutual coupling effect, establishing a mutual coupling matrix, and establishing a signal model according to the mutual coupling matrix and the array signal expression, the signal model characterizing the array signal received in the presence of a mutual coupling effect.

[0063] Alternatively, for a non-uniform linear array, the irregularity of the array element positions makes it necessary to consider the geometric relationship of each array element relative to the reference point (or reference array element) when determining the DOA expression of the source signal. Assume that the non-uniform linear array is located on the x-axis and the array element position is x n (n=1,2,...,N, N is the number of array elements), the array operating wavelength is λ, and for a far-field source signal, its arrival direction is θ (relative to the normal direction of the linear array). According to the wave propagation principle, the time delay of the signal reaching different array elements is related to the array element position and the signal arrival direction. The phase difference of the signal received by the array element relative to the reference array element is for , where x0 is the position of the reference array element. This expression is the DOA expression corresponding to a single source signal, which describes the relationship between the arrival direction of the source signal and the phase difference of the signal received by the array element. When there are multiple source signals, assume that there are M source signals, and their arrival directions are In the absence of mutual coupling effect, according to the linear superposition principle, the signal x received by the nth array element is n (t) is the superposition of various source signals on the array element.

[0064] Exemplarily, as shown in FIG3 , K unrelated source signals From the far field incident on the spatial M-element non-uniform linear array, where the reference element is at the origin of the coordinate system, and the distance between the mth element and the reference element is d m , the array working wavelength is λ, represents the DOA of the kth source signal, k=1,2,...,K.

[0065] The array signal received without mutual coupling is:

[0066]

[0067] in, is the received signal vector of all array elements, x m (t) represents the received signal of the mth array element at time t, and the superscript T represents the transposition. is the response vector of the Kth unrelated source, is the kth baseband signal, is the array noise is the direction matrix, is the source signal matrix; The specific form is expressed as:

[0068]

[0069] Assume that the noise is additive Gaussian white noise, that is:

[0070]

[0071] in, It means to find the mathematical expectation, I M represents the identity matrix of dimension M×M, is the noise variance.

[0072] When mutual coupling effect exists between array elements, an M×M mutual coupling matrix C is introduced to describe the mutual coupling effect between M array elements, where the reference array element is not affected by the mutual coupling. The specific form of the mutual coupling matrix is ​​as follows:

[0073]

[0074] Among them, c located at (p,q) in C m is the mutual coupling coefficient between the pth array element and the qth array element; under ideal conditions, c m The size of is inversely proportional to the distance between the array elements. If the distance between the two sets of array elements is the same, the mutual coupling coefficient is also the same, and for any m>1, .

[0075] Then the array signal in the presence of mutual coupling can be rewritten as:

[0076]

[0077] Among them, if is Gaussian white noise and is similar to the source signal Not relevant, then The covariance matrix of is:

[0078]

[0079] in, Source signal The covariance matrix of represents a diagonal matrix whose diagonal elements are the numbers in brackets, is the power of the Kth source signal, is the noise variance; when the number of snapshots L is given, let , the covariance matrix R is estimated by:

[0080]

[0081] The superscript H represents the conjugate transpose.

[0082] In this embodiment, by considering the signal model of mutual coupling effect in array signal processing, the signal received by the array can be more realistically reflected. Furthermore, by adopting the signal model considering the mutual coupling effect, the array manifold can be constructed more accurately, thereby improving the accuracy of DOA estimation.

[0083] In an exemplary embodiment, the step of dividing the non-uniform linear array into a plurality of sub-arrays according to the spacing relationship between array elements includes: determining a mutual coupling range according to an operating wavelength of the non-uniform linear array; and dividing the non-uniform linear array into a plurality of sub-arrays according to the mutual coupling range, wherein there is no mutual coupling effect between array elements in each sub-array.

[0084] Optionally, in a non-uniform linear array, the mutual coupling effect between array elements is closely related to the working wavelength, and the mutual coupling effect is caused by the electromagnetic field interaction between array elements. When the array element spacing is small, this interaction will be more obvious relative to the working wavelength. When the array element spacing is less than a certain multiple (for example, less than half the working wavelength), the mutual coupling effect cannot be ignored. This multiple can be determined based on the specific electromagnetic field theory and the characteristics of the antenna array. According to this principle, a distance range can be determined based on the working wavelength of the non-uniform linear array. There is an obvious mutual coupling effect between the array elements within this range. This range is the mutual coupling range, thereby dividing the linear array into N sub-arrays. The array elements within the sub-arrays are allowed to overlap, but there must be no mutual coupling effect.

[0085] Exemplarily, as shown in FIG3 , there are antenna array elements composed of M=10 sensors, the working wavelength is 1, the antenna numbers are 1, 2, ..., 10, and the position coordinates of the sensors are [0, 3.42, 3.59, 5.24, 6.00, 7.47, 7.66, 8.50, 10.10, 10.73, respectively. If the array element spacing is less than or equal to 2, the antenna has a mutual coupling effect. All array elements are divided into 3 groups of subarrays. For example, the first group of subarrays includes antenna array elements numbered 1, 2, 5, 8, and 10, the second group of subarrays includes antenna array elements numbered 1, 3, 6, and 9, and the third group of subarrays includes antenna array elements numbered 1, 4, and 7.

[0086] Furthermore, when sampling the subarrays, at the first sampling moment, the first group of subarray signals are sampled, and the antennas of the remaining groups are in a "power-off" state; at the second sampling moment, the second group of subarray element signals are sampled, and so on, until the Nth group of subarray signals are collected; the first group to the Nth group of subarray signals are resampled in sequence; taking the result of the above three subarray divisions as an example, the specific sampling is as follows:

[0087] For example, at the first sampling time t1, only the signals of the first group of array elements are synchronously sampled, and the remaining groups of antennas are in the "power-off" state, and the obtained , , , ; At the second sampling time t2, only the signals of the second group of array elements are synchronously sampled, and the remaining groups of antennas are in the "power-off" state, and we get , , , ; At the third sampling moment, only the signal of the third group of array elements is sampled, and the antennas of the other groups are in the "power-off" state; at the fourth sampling moment, only the signal of the first group of antennas is sampled, and the antennas of the other groups are in the "power-off" state, and so on; the final sampling data It can be expressed as:

[0088]

[0089] In this embodiment, by determining the mutual coupling range according to the working wavelength of the non-uniform linear array, the non-uniform linear array is divided into a plurality of sub-arrays according to the mutual coupling range, and there is no mutual coupling effect between the array elements in each sub-array, so that the influence of the mutual coupling effect on the response of the array to signals in different directions can be eliminated, thereby improving the accuracy of DOA estimation.

[0090] In an exemplary embodiment, the step of approximating each element in the signal covariance matrix to obtain an approximate covariance matrix includes: for each element in the signal covariance matrix, decomposing the element by a Bessel function to obtain a Bessel combination expression corresponding to the element; determining the number of approximation terms corresponding to the element, converting the Bessel combination expression into an approximate expression according to the number of approximation terms, and obtaining a corresponding approximate covariance matrix.

[0091] Optionally, in signal processing, the elements of the signal covariance matrix reflect characteristics such as correlation between signals. For each element in the signal covariance matrix, it is decomposed by the Bessel function. For example, when considering the mutual coupling effect between array elements and the propagation mode of the signal in space, the Bessel function can well describe these complex relationships, and the Bessel combination expression corresponding to the obtained element can more accurately reflect the physical meaning and mathematical relationship contained in the element.

[0092] Furthermore, in the process of converting the Bessel combinatorial expression into an approximate expression, determining the number of approximation terms is a key step. The determination of the number of approximation terms usually needs to consider multiple factors, such as the required approximation accuracy, the limitation of computing resources, etc. If a higher approximation accuracy is required, it may be necessary to select more approximation terms, but this will increase the computational complexity. On the contrary, if the computing resources are limited or the accuracy requirement is not very high, a smaller number of approximation terms can be selected. The process of determining the number of approximation terms may need to be weighed according to the specific application scenario and system requirements. Afterwards, the Bessel combinatorial expression is converted into an approximate expression according to the determined number of approximation terms. This process is actually an operation of truncating or simplifying the Bessel combinatorial expression. For example, if the Bessel combinatorial expression is in the form of an infinite series, the first several terms are retained according to the number of approximation terms, while the latter terms are ignored, thereby obtaining an approximate expression. The approximate expression obtained in this way can greatly simplify the calculation process while meeting certain accuracy requirements, and finally obtain the corresponding approximate covariance matrix.

[0093] Exemplarily, the signal covariance matrix estimate calculated based on the collected sub-array signals is It can be expressed as:

[0094]

[0095] in, , the superscript * indicates the conjugate operation. Since the signal covariance matrix has missing data at this time, it is impossible to directly decompose the sub-controls. Therefore, the convex optimization method is used to complete it. The missing elements in .

[0096] For example, let express The mth element of It can be broken down into:

[0097]

[0098] In the above formula, represents the nth order Bessel function of the first kind. In fact, since the Bessel function converges rapidly with n, it can be approximated by a finite Bessel function In engineering, N coefficients are often chosen for approximation, where N needs to satisfy: .

[0099] at this time, can be approximated as:

[0100]

[0101] Therefore,

[0102]

[0103] in,

[0104]

[0105]

[0106] Under ideal conditions without mutual coupling, the covariance matrix of the received signal is:

[0107]

[0108] in, , Obviously, It has a conjugate symmetric Toeplitz structure, that is

[0109]

[0110] In the above formula, u is a vector with a dimension of 2N+1. Returns a Toeplitz matrix whose first column is u.

[0111] In this embodiment, for each element in the signal covariance matrix, the element is decomposed by the Bessel function to obtain the Bessel combination expression corresponding to the element, the number of approximation terms corresponding to the element is determined, and the Bessel combination expression is converted into an approximate expression according to the number of approximation terms to obtain the corresponding approximate covariance matrix. This can reduce the computational complexity while retaining the key information in the signal covariance matrix to a certain extent, thereby improving the accuracy of subsequent DOA estimation.

[0112] In an exemplary embodiment, the step of solving the approximate covariance matrix according to all index positions to obtain a target covariance matrix includes: establishing an objective function according to the matrix structure of the approximate covariance matrix, the objective function including the trace of the matrix, a regularization parameter and the norm of the matrix; determining constraints according to the index positions; and solving the objective function under the constraints to obtain the target covariance matrix.

[0113] Optionally, the trace of a matrix is ​​the sum of the elements on the main diagonal of the matrix. In signal processing, the trace of a covariance matrix is ​​often related to the total power of the signal. For example, for a signal covariance matrix, its trace represents the sum of the powers of all source signals (in the ideal case of noise-free). Including the trace of the matrix when constructing the objective function can reflect information related to the power characteristics of the signal. The regularization parameter is used to control the weights of different terms in the objective function, and can balance the relationship between the trace of the matrix and the norm of the matrix. The norm of the matrix can measure the size or "length" of the matrix. Including the norm of the matrix in the objective function can constrain the structure of the approximate covariance matrix, such as limiting the size range of the matrix elements to avoid elements that are too large or too small, so that the resulting target covariance matrix has a more reasonable structure.

[0114] Optionally, determining constraints based on index positions means defining constraints based on elements at specific positions in the approximate covariance matrix or substructures associated with these indices. For example, if the index positions correspond to specific source signals or specific subarray relationships, then the constraints may restrict the elements at these positions to satisfy certain relationships.

[0115] Furthermore, after determining the objective function and constraints, the objective function can be solved to obtain the target covariance matrix. Among them, the advantage of the convex optimization algorithm is that it can guarantee the finding of the global optimal solution (in the case of convex problems). For example, if the objective function is a convex function and the feasible domain formed by the constraints is a convex set, then a convex optimization algorithm such as the interior point method can be used to solve the objective function. In this way, while satisfying the constraints, the matrix that makes the objective function reach the optimal value is found, that is, the target covariance matrix.

[0116] For example, let Represents an index collection whose records The index position of the non-zero elements in . u can be solved by the following convex optimization method:

[0117]

[0118] in, represents the trace of the matrix, is a regularization parameter, represents the F-norm of the matrix, means the matrix is ​​positive definite, express The index is Finally, the convex optimization toolbox can be used directly to solve the above problem.

[0119] In this embodiment, by establishing an objective function and considering the trace of the matrix, the regularization parameter and the norm of the matrix, the structure of the target covariance matrix can be optimized, so that the matrix is ​​more stable in subsequent signal processing operations, and in the case of multi-source signals, the optimized matrix structure can better reflect the relationship between different source signals and improve the accuracy of the description of signal characteristics.

[0120] In an exemplary embodiment, the target covariance matrix is ​​eigen-decomposed, and DOA estimation is performed based on the result of the eigen-decomposition, including: performing eigen-decomposition on the target covariance matrix to obtain signal subspace characteristics and noise subspace characteristics; establishing a spatial spectrum function based on the signal subspace characteristics and the noise subspace characteristics; and determining the DOA of the source signal based on the peak value of the spatial spectrum function.

[0121] Optionally, the target covariance matrix is ​​a symmetric matrix. By performing eigendecomposition on the target covariance matrix, it can be decomposed into signal subspace features and noise subspace features. For example, in the case of a source signal, the eigenvectors corresponding to the first larger eigenvalues ​​constitute the signal subspace features, and the eigenvectors corresponding to the remaining eigenvalues ​​constitute the noise subspace features. Afterwards, a spatial spectrum function is constructed based on the signal subspace features and the noise subspace features. The common method of constructing a spatial spectrum function is based on the orthogonality principle of the signal subspace and the noise subspace. For example, in the MUSIC (Multiple Signal Classification) algorithm, the spatial spectrum function It can be expressed as ,in, is the array manifold vector, which is related to the direction of arrival of the signal Correlation, indicating that the array is from the direction response to the signal.

[0122] Furthermore, after the spatial spectrum function is calculated, the DOA of the source signal is determined based on its peak value. Because the value of the spatial spectrum function represents the possibility of the signal existing in different directions, the direction where the peak appears is the most likely direction of arrival of the signal. For example, by searching the spatial spectrum function within a certain angle range, The maximum value of The maximum value is obtained at , then the arrival direction of the source signal can be determined as .

[0123] Exemplarily, as shown in FIG5 , assume that there are K=3 far-field random source signals, and their DOAs are 10°, 20°, and 60°, respectively. Assume that there are M=32 randomly distributed linear arrays, and the wavelength of the antenna operation is 1. If the array element spacing is less than 2, there will be mutual coupling effects. According to the proposed grouping scheme, the array elements are grouped and sampled. Under the conditions that the number of sampling samples is L=300, the number of sampling samples in each group is 100, and the signal-to-noise ratio is 10dB, the MUSIC method is used for angle estimation. The scatter plot results of 100 Monte Carlo simulations are shown in FIG5 . It can be seen from FIG5 that for any array manifold and mutual coupling effects related to the angle of arrival, the mutual coupling effects can be suppressed by space-time mutually exclusive sampling, so that the DOA of the source signal can be effectively obtained using the existing DOA algorithm.

[0124] In this embodiment, the target covariance matrix is ​​subjected to eigendecomposition to obtain signal subspace characteristics and noise subspace characteristics, and a spatial spectrum function is established based on the signal subspace characteristics and the noise subspace characteristics. The DOA of the source signal is determined according to the peak value of the spatial spectrum function, and the signal subspaces corresponding to different source signals can be distinguished, and then a suitable spatial spectrum function is constructed to estimate the DOA of each source signal, thereby improving the accuracy of DOA estimation.

[0125] In an exemplary embodiment, as shown in FIG6 , a DOA estimation method based on a non-uniform linear array is provided, and the method includes the following steps:

[0126] According to the element positions and array operating wavelength of the non-uniform linear array, the DOA expression corresponding to each source signal is determined, and the array signal expression is determined according to the DOA expressions corresponding to all source signals. The signal expression is used to characterize the ideal array signal received in the absence of mutual coupling effect. In the presence of mutual coupling effect, a mutual coupling matrix is ​​established, and a signal model is established based on the mutual coupling matrix and the array signal expression. The signal model characterizes the array signal received in the presence of mutual coupling effect.

[0127] The mutual coupling range is determined according to the working wavelength of the non-uniform linear array; the non-uniform linear array is divided into a plurality of sub-arrays according to the mutual coupling range, and there is no mutual coupling effect between the array elements in each sub-array.

[0128] According to the preset sampling order, the subarray signals of the target subarray corresponding to the sampling order are collected in sequence; the target subarray corresponding to each sampling order is in a normal working state, and the subarrays other than the target subarray are in a non-working state.

[0129] A signal covariance matrix is ​​established according to the signal model and all sub-array signals. For each element in the signal covariance matrix, the element is decomposed by the Bessel function to obtain the Bessel combination expression corresponding to the element; the number of approximation terms corresponding to the element is determined, and the Bessel combination expression is converted into an approximate expression according to the number of approximation terms to obtain the corresponding approximate covariance matrix.

[0130] Determine the index position of the approximate covariance matrix, establish the objective function according to the matrix structure of the approximate covariance matrix, the objective function includes the trace of the matrix, the regularization parameter and the norm of the matrix; determine the constraint conditions according to the index position; solve the objective function under the constraint conditions to obtain the target covariance matrix.

[0131] The target covariance matrix is ​​eigen-decomposed to obtain the signal subspace characteristics and the noise subspace characteristics; based on the signal subspace characteristics and the noise subspace characteristics, a spatial spectrum function is established; and according to the peak value of the spatial spectrum function, the DOA of the source signal is determined.

[0132] In this embodiment, by establishing a signal model of a source signal and a non-uniform linear array, the non-uniform linear array is divided into a plurality of sub-arrays according to the spacing relationship between array elements, a signal covariance matrix is ​​established according to the signal model and all sub-array signals, and each element in the signal covariance matrix is ​​approximated to obtain an approximate covariance matrix, which can suppress the influence of noise on signal processing to a certain extent. When DOA estimation is finally performed, the direction information of the signal can be more accurately extracted from the noise background, thereby improving the accuracy of DOA estimation. In addition, by sequentially collecting the sub-array signals of the target sub-array corresponding to the sampling sequence according to a preset sampling sequence, only part of the sub-array works each time, which can reduce the number of hardware devices working simultaneously, thereby reducing the hardware complexity. At the same time, through simple space-time mutually exclusive sampling, the mutual coupling effect can be suppressed, thereby effectively reducing the complexity of the operation process and improving the response speed.

[0133] It should be understood that, although the steps in the flowcharts involved in the above embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0134] Based on the same inventive concept, the embodiment of the present application also provides a DOA estimation device based on a non-uniform linear array for implementing the DOA estimation method based on a non-uniform linear array. The implementation solution provided by the device to solve the problem is similar to the implementation solution recorded in the above method, so the specific limitations in one or more embodiments of the DOA estimation device based on a non-uniform linear array provided below can refer to the limitations of the DOA estimation method based on a non-uniform linear array above, and will not be repeated here.

[0135] In an exemplary embodiment, as shown in FIG7 , a DOA estimation device based on a non-uniform linear array is provided, comprising:

[0136] The model building module 10 is used to build a signal model of a source signal and a non-uniform linear array.

[0137] The linear array division module 20 is used to divide the non-uniform linear array into multiple sub-arrays according to the spacing relationship between array elements, and collect sub-array signals of the target sub-array corresponding to the sampling sequence in sequence according to a preset sampling sequence; the target sub-array corresponding to each sampling sequence is in a normal working state, and the sub-arrays other than the target sub-array are in a non-working state.

[0138] The matrix calculation module 30 is used to establish a signal covariance matrix according to the signal model and all sub-array signals, and perform approximate processing on each element in the signal covariance matrix to obtain an approximate covariance matrix.

[0139] The matrix solving module 40 is used to determine the index position of the approximate covariance matrix, solve the approximate covariance matrix according to all the index positions, and obtain the target covariance matrix.

[0140] The DOA estimation module 50 is used to perform eigendecomposition on the target covariance matrix and perform DOA estimation based on the result of the eigendecomposition.

[0141] In an exemplary embodiment, the model building module 10 is also used to determine the DOA expression corresponding to each source signal according to the array element position and the array operating wavelength of the non-uniform linear array, and determine the array signal expression according to the DOA expressions corresponding to all source signals, and the signal expression is used to characterize the ideal array signal received in the absence of mutual coupling effect; in the presence of mutual coupling effect, a mutual coupling matrix is ​​established, and a signal model is established according to the mutual coupling matrix and the array signal expression, and the signal model characterizes the array signal received in the presence of mutual coupling effect.

[0142] In an exemplary embodiment, the linear array division module 20 is further used to determine the mutual coupling range according to the working wavelength of the non-uniform linear array; the non-uniform linear array is divided into multiple sub-arrays according to the mutual coupling range, and there is no mutual coupling effect between the array elements in each sub-array.

[0143] In an exemplary embodiment, the matrix calculation module 30 is also used to decompose each element in the signal covariance matrix through the Bessel function to obtain the Bessel combination expression corresponding to the element; determine the number of approximation terms corresponding to the element, convert the Bessel combination expression into an approximate expression according to the number of approximate terms, and obtain the corresponding approximate covariance matrix.

[0144] In an exemplary embodiment, the matrix solving module 40 is also used to establish an objective function based on the matrix structure of the approximate covariance matrix, the objective function including the trace of the matrix, the regularization parameter and the norm of the matrix; determine the constraint conditions according to the index position; under the constraint conditions, solve the objective function to obtain the target covariance matrix.

[0145] In an exemplary embodiment, the DOA estimation module 50 is also used to perform eigendecomposition on the target covariance matrix to obtain signal subspace characteristics and noise subspace characteristics; establish a spatial spectrum function based on the signal subspace characteristics and the noise subspace characteristics; and determine the DOA of the source signal according to the peak value of the spatial spectrum function.

[0146] Each module in the above-mentioned DOA estimation device based on non-uniform linear array can be implemented in whole or in part by software, hardware and a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0147] In an exemplary embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be shown in FIG8. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit, and an input device. The processor, the memory, and the input / output interface are connected via a system bus, and the communication interface, the display unit, and the input device are connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner may be implemented through WIFI, a mobile cellular network, near field communication (NFC) or other technologies. When the computer program is executed by the processor, a DOA estimation method based on a non-uniform linear array is implemented. The display unit of the computer device is used to form a visually visible picture, which may be a display screen, a projection device, or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0148] Those skilled in the art will understand that the structure shown in FIG. 8 is merely a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0149] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the following steps when executing the computer program: establishing a signal model of a source signal and a non-uniform linear array; dividing the non-uniform linear array into a plurality of sub-arrays according to the spacing relationship between array elements, and sequentially collecting sub-array signals of a target sub-array corresponding to a sampling sequence according to a preset sampling sequence; the target sub-array corresponding to each sampling sequence is in a normal working state, and other sub-arrays except the target sub-array are in a non-working state; establishing a signal covariance matrix according to the signal model and all sub-array signals, approximating each element in the signal covariance matrix to obtain an approximate covariance matrix; determining the index position of the approximate covariance matrix, solving the approximate covariance matrix according to all index positions to obtain a target covariance matrix; eigendecomposing the target covariance matrix, and performing DOA estimation based on the result of the eigendecomposition.

[0150] In one embodiment, the establishment of a signal model of a source signal and a non-uniform linear array involved in the execution of a computer program by a processor includes: determining a DOA expression corresponding to each source signal according to an array element position and an array operating wavelength of the non-uniform linear array, and determining an array signal expression according to the DOA expressions corresponding to all source signals, the signal expression being used to characterize an ideal array signal received in the absence of a mutual coupling effect; in the presence of a mutual coupling effect, establishing a mutual coupling matrix, and establishing a signal model according to the mutual coupling matrix and the array signal expression, the signal model characterizing the array signal received in the presence of a mutual coupling effect.

[0151] In one embodiment, the method of dividing a non-uniform linear array into a plurality of sub-arrays according to a spacing relationship between array elements when a processor executes a computer program includes: determining a mutual coupling range according to an operating wavelength of the non-uniform linear array; and dividing the non-uniform linear array into a plurality of sub-arrays according to the mutual coupling range, wherein there is no mutual coupling effect between array elements in each sub-array.

[0152] In one embodiment, the processor executes a computer program that involves approximating each element in a signal covariance matrix to obtain an approximate covariance matrix, including: for each element in the signal covariance matrix, decomposing the element by a Bessel function to obtain a Bessel combination expression corresponding to the element; determining the number of approximation terms corresponding to the element, converting the Bessel combination expression into an approximate expression according to the number of approximation terms, and obtaining a corresponding approximate covariance matrix.

[0153] In one embodiment, when a processor executes a computer program, the processor solves the approximate covariance matrix according to all index positions to obtain a target covariance matrix, including: establishing an objective function according to the matrix structure of the approximate covariance matrix, the objective function including the trace of the matrix, a regularization parameter and the norm of the matrix; determining constraints according to the index positions; and solving the objective function under the constraints to obtain the target covariance matrix.

[0154] In one embodiment, when a processor executes a computer program, it involves performing eigendecomposition on a target covariance matrix and performing DOA estimation based on the result of the eigendecomposition, including: performing eigendecomposition on the target covariance matrix to obtain signal subspace characteristics and noise subspace characteristics; establishing a spatial spectrum function based on the signal subspace characteristics and the noise subspace characteristics; and determining the DOA of the source signal based on the peak value of the spatial spectrum function.

[0155] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented: establishing a signal model of a source signal and a non-uniform linear array; dividing the non-uniform linear array into a plurality of sub-arrays according to the spacing relationship between array elements, and sequentially acquiring sub-array signals of a target sub-array corresponding to a sampling sequence according to a preset sampling sequence; the target sub-array corresponding to each sampling sequence is in a normal working state, and the sub-arrays other than the target sub-array are in a non-working state; establishing a signal covariance matrix according to the signal model and all sub-array signals, approximating each element in the signal covariance matrix to obtain an approximate covariance matrix; determining the index position of the approximate covariance matrix, solving the approximate covariance matrix according to all index positions to obtain a target covariance matrix; eigendecomposing the target covariance matrix, and performing DOA estimation based on the result of the eigendecomposition.

[0156] In one embodiment, the computer program involved in establishing a signal model of a source signal and a non-uniform linear array when executed by a processor includes: determining a DOA expression corresponding to each source signal according to an array element position and an array operating wavelength of the non-uniform linear array, and determining an array signal expression according to the DOA expressions corresponding to all source signals, the signal expression being used to characterize an ideal array signal received in the absence of a mutual coupling effect; in the presence of a mutual coupling effect, establishing a mutual coupling matrix, and establishing a signal model according to the mutual coupling matrix and the array signal expression, the signal model characterizing the array signal received in the presence of a mutual coupling effect.

[0157] In one embodiment, the computer program executed by the processor involves dividing the non-uniform linear array into a plurality of sub-arrays according to the spacing relationship between array elements, including: determining a mutual coupling range according to an operating wavelength of the non-uniform linear array; and dividing the non-uniform linear array into a plurality of sub-arrays according to the mutual coupling range, wherein there is no mutual coupling effect between array elements in each sub-array.

[0158] In one embodiment, when a computer program is executed by a processor, the computer program involves approximating each element in a signal covariance matrix to obtain an approximate covariance matrix, including: for each element in the signal covariance matrix, decomposing the element by a Bessel function to obtain a Bessel combination expression corresponding to the element; determining the number of approximation terms corresponding to the element, converting the Bessel combination expression into an approximate expression according to the number of approximation terms, and obtaining a corresponding approximate covariance matrix.

[0159] In one embodiment, when the computer program is executed by a processor, the computer program involves solving the approximate covariance matrix according to all index positions to obtain a target covariance matrix, including: establishing an objective function according to the matrix structure of the approximate covariance matrix, the objective function including the trace of the matrix, a regularization parameter and the norm of the matrix; determining constraints according to the index positions; and solving the objective function under the constraints to obtain the target covariance matrix.

[0160] In one embodiment, when a computer program is executed by a processor, it involves performing eigendecomposition on a target covariance matrix and performing DOA estimation based on the result of the eigendecomposition, including: performing eigendecomposition on the target covariance matrix to obtain signal subspace characteristics and noise subspace characteristics; establishing a spatial spectrum function based on the signal subspace characteristics and the noise subspace characteristics; and determining the DOA of the source signal based on the peak value of the spatial spectrum function.

[0161] In one embodiment, a computer program product is provided, including a computer program, which implements the following steps when executed by a processor: establishing a signal model of a source signal and a non-uniform linear array; dividing the non-uniform linear array into a plurality of sub-arrays according to the spacing relationship between array elements, and sequentially acquiring sub-array signals of a target sub-array corresponding to a sampling sequence according to a preset sampling sequence; the target sub-array corresponding to each sampling sequence is in a normal working state, and the sub-arrays other than the target sub-array are in a non-working state; establishing a signal covariance matrix according to the signal model and all sub-array signals, approximating each element in the signal covariance matrix to obtain an approximate covariance matrix; determining the index position of the approximate covariance matrix, solving the approximate covariance matrix according to all index positions to obtain a target covariance matrix; eigendecomposing the target covariance matrix, and performing DOA estimation based on the result of the eigendecomposition.

[0162] In one embodiment, the computer program involved in establishing a signal model of a source signal and a non-uniform linear array when executed by a processor includes: determining a DOA expression corresponding to each source signal according to an array element position and an array operating wavelength of the non-uniform linear array, and determining an array signal expression according to the DOA expressions corresponding to all source signals, the signal expression being used to characterize an ideal array signal received in the absence of a mutual coupling effect; in the presence of a mutual coupling effect, establishing a mutual coupling matrix, and establishing a signal model according to the mutual coupling matrix and the array signal expression, the signal model characterizing the array signal received in the presence of a mutual coupling effect.

[0163] In one embodiment, the computer program executed by the processor involves dividing the non-uniform linear array into a plurality of sub-arrays according to the spacing relationship between array elements, including: determining a mutual coupling range according to an operating wavelength of the non-uniform linear array; and dividing the non-uniform linear array into a plurality of sub-arrays according to the mutual coupling range, wherein there is no mutual coupling effect between array elements in each sub-array.

[0164] In one embodiment, when a computer program is executed by a processor, the computer program involves approximating each element in a signal covariance matrix to obtain an approximate covariance matrix, including: for each element in the signal covariance matrix, decomposing the element by a Bessel function to obtain a Bessel combination expression corresponding to the element; determining the number of approximation terms corresponding to the element, converting the Bessel combination expression into an approximate expression according to the number of approximation terms, and obtaining a corresponding approximate covariance matrix.

[0165] In one embodiment, when the computer program is executed by a processor, the computer program involves solving the approximate covariance matrix according to all index positions to obtain a target covariance matrix, including: establishing an objective function according to the matrix structure of the approximate covariance matrix, the objective function including the trace of the matrix, a regularization parameter and the norm of the matrix; determining constraints according to the index positions; and solving the objective function under the constraints to obtain the target covariance matrix.

[0166] In one embodiment, when a computer program is executed by a processor, it involves performing eigendecomposition on a target covariance matrix and performing DOA estimation based on the result of the eigendecomposition, including: performing eigendecomposition on the target covariance matrix to obtain signal subspace characteristics and noise subspace characteristics; establishing a spatial spectrum function based on the signal subspace characteristics and the noise subspace characteristics; and determining the DOA of the source signal based on the peak value of the spatial spectrum function.

[0167] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0168] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0169] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A DOA estimation method based on a non-uniform linear array, characterized in that: The method comprises: Establish the signal model of source signal and non-uniform linear array; According to the spacing relationship between array elements, the non-uniform linear array is divided into a plurality of sub-arrays, and sub-array signals of the target sub-arrays corresponding to the sampling order are sequentially collected according to a preset sampling order; each target sub-array corresponding to the sampling order is in a normal working state, and other sub-arrays except the target sub-array are in a non-working state; Establishing a signal covariance matrix according to the signal model and all sub-array signals, and performing approximate processing on each element in the signal covariance matrix to obtain an approximate covariance matrix; Determine the index position of the approximate covariance matrix, and solve the approximate covariance matrix according to all the index positions to obtain a target covariance matrix; The target covariance matrix is ​​subjected to eigendecomposition, and DOA estimation is performed based on the result of the eigendecomposition.

2. The method according to claim 1, characterized in that: The step of establishing a signal model of a source signal and a non-uniform linear array includes: According to the array element positions and the array operating wavelength of the non-uniform linear array, the DOA expression corresponding to each source signal is determined, and the array signal expression is determined according to the DOA expressions corresponding to all source signals, wherein the signal expression is used to characterize the ideal array signal received in the absence of mutual coupling effect; In the case of the existence of mutual coupling effect, a mutual coupling matrix is ​​established, and a signal model is established according to the mutual coupling matrix and the array signal expression, wherein the signal model represents the array signal received in the case of the existence of the mutual coupling effect.

3. The method according to claim 1, characterized in that The step of dividing the non-uniform linear array into a plurality of sub-arrays according to the spacing relationship between array elements comprises: Determining a mutual coupling range according to the operating wavelength of the non-uniform linear array; The non-uniform linear array is divided into a plurality of sub-arrays according to the mutual coupling range, and there is no mutual coupling effect between array elements in each of the sub-arrays.

4. The method according to claim 1, characterized in that: The step of performing approximate processing on each element in the signal covariance matrix to obtain an approximate covariance matrix includes: For each element in the signal covariance matrix, decompose the element by using a Bessel function to obtain a Bessel combination expression corresponding to the element; The number of approximation terms corresponding to the element is determined, and the Bessel combination expression is converted into an approximate expression according to the number of approximation terms to obtain a corresponding approximate covariance matrix.

5. The method according to claim 1, characterized in that The method of solving the approximate covariance matrix according to all index positions to obtain a target covariance matrix includes: Establishing an objective function according to the matrix structure of the approximate covariance matrix, wherein the objective function includes a trace of the matrix, a regularization parameter, and a norm of the matrix; Determining a constraint condition according to the index position; Under the constraint conditions, the objective function is solved to obtain a target covariance matrix.

6. The method according to claim 1, characterized in that The performing eigendecomposition on the target covariance matrix and performing DOA estimation based on the result of the eigendecomposition includes: Performing eigendecomposition on the target covariance matrix to obtain signal subspace features and noise subspace features; Establishing a spatial spectrum function based on the signal subspace characteristics and the noise subspace characteristics; The DOA of the source signal is determined according to the peak value of the spatial spectrum function.

7. A DOA estimation device based on a non-uniform linear array, characterized in that: The device comprises: A model building module, used to build a signal model of a source signal and a non-uniform linear array; A linear array division module, used to divide the non-uniform linear array into a plurality of sub-arrays according to the spacing relationship between array elements, and sequentially collect sub-array signals of target sub-arrays corresponding to the sampling order according to a preset sampling order; each target sub-array corresponding to the sampling order is in a normal working state, and other sub-arrays except the target sub-array are in a non-working state; A matrix calculation module, used for establishing a signal covariance matrix according to the signal model and all sub-array signals, and performing approximate processing on each element in the signal covariance matrix to obtain an approximate covariance matrix; A matrix solving module, used for determining the index position of the approximate covariance matrix, solving the approximate covariance matrix according to all the index positions, and obtaining a target covariance matrix; The DOA estimation module is used to perform eigendecomposition on the target covariance matrix and perform DOA estimation based on the result of the eigendecomposition.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images