A method and system for estimating the direction of arrival of a two-dimensional coherent signal source in a strong impulsive noise environment

Abstract

The application discloses a two-dimensional coherent signal source direction of arrival estimation method and system under strong impulse noise environment, relates to the field of array signal processing technology, and aims at solving the problem of impulse noise and coherent signal direction finding in a severe electromagnetic environment which cannot be effectively processed by the existing method. The technical points of the application include the following steps: step one, establishing a two-dimensional sampling signal model of a uniform surface array under impulse noise environment, and obtaining the snapshot data of the surface array; step two, constructing a covariance matrix based on an infinite norm Cauchy kernel correlation entropy, and obtaining a weighted subspace fitting equation; step three, constructing an adaptability function according to the fitting equation, and initializing a quantum wild swallow population; step four, updating the quantum position of the quantum wild swallow according to a migration strategy; step five, updating the quantum position of the quantum wild swallow according to a breeding strategy; step six, updating the global optimal quantum position; step seven, updating the population size according to the death strategy of the quantum wild swallow; and finally, the estimated direction of the incoming wave signal is obtained.

Description

Technical Field

[0001] This invention relates to the field of array signal processing technology, and more specifically, to a method and system for estimating the direction of arrival of a two-dimensional coherent signal source under strong impulse noise conditions. Background Technology

[0002] Direction of arrival (DOA) estimation is a technique based on antenna arrays to determine the direction of arrival (DOA), and it has been widely used in wireless communication, radio astronomy, and radar. Compared to one-dimensional DOA estimation, which can only estimate angular parameters within a plane, two-dimensional DOA estimation requires simultaneous estimation of the elevation and azimuth angles of the incoming wave, making it better suited to the direction estimation needs of complex space scenarios. However, its requirement to jointly solve for the elevation and azimuth angles of the incoming wave also results in a significant increase in computational load. Currently, the main two-dimensional DOA estimation methods are the two-dimensional MUSIC algorithm and the two-dimensional ESPRIT algorithm and their improved algorithms, which are based on two-dimensional spatial spectra. The two-dimensional MUSIC algorithm requires spectral peak search in two-dimensional space, which has a large computational load and limited estimation accuracy. The two-dimensional ESPRIT algorithm avoids the spectral peak search process, but it requires additional angle pairing operations and is prone to pairing errors when there are many incident sources, thus affecting the estimation results. There is also the PM algorithm proposed under uniform array conditions, which constructs a propagation operator through linear operations to estimate the direction of arrival. It can not only reduce the computational complexity but also realize automatic pairing of two-dimensional parameters. However, this algorithm performs poorly in harsh environments such as impulse noise and requires additional decoherence operations to process coherent signals.

[0003] Based on existing technical literature, Tian Zhengdong et al.'s "Two-Dimensional DOA Estimation Algorithm Based on Propagation Operator in Area Arrays," published in *Computer Engineering and Applications*, proposes a PM-based estimation algorithm for uniform area arrays. This algorithm reuses data in the cross-correlation matrix by utilizing the cross-correlation information between subarrays divided by the uniform area array, and obtains rotation-invariant relations by constructing a propagation operator to achieve two-dimensional DOA estimation. It avoids the eigenvalue decomposition process of subspace algorithms, achieving high-precision estimation while automatically matching parameters. However, this algorithm is only applicable to ideal Gaussian noise and cannot handle coherent signals. Zhang Guohui et al.'s "Simulation and Implementation of Subarray Spatial Smoothing Algorithm Based on Uniform Area Array," published in *Computer Simulation*, proposes a method to construct the covariance matrix after dividing a uniform area array into subarrays based on spatial smoothing technology. This achieves 360-degree omnidirectional direction finding while effectively processing coherent signals, reducing the computational load to some extent. However, this method is still limited to ideal Gaussian noise environments and cannot achieve DOA estimation under impulse noise environments. Moreover, spatial smoothing technology usually causes aperture loss in the array. The paper "Two-dimensional DOA estimation under impulse noise environment" published by Li Miaomiao et al. in "Aerospace Defense" successfully achieved two-dimensional DOA estimation under impulse noise environment by constructing a new correlation entropy covariance matrix to replace the autocorrelation function in the traditional subspace algorithm and combining it with the ESPRIT method on the basis of a uniform circular array. However, this method does not have the ability to process coherent signals and requires additional operations to achieve decoherence.

[0004] Existing research indicates that current two-dimensional DOA estimation algorithms are mainly based on subspace decomposition or propagation operators constructed based on uniform array structures. However, in practical applications, due to electromagnetic interference and multipath effects, the signals arriving at the base station antenna array are usually coherent and easily affected by impulse noise. Currently, there is a lack of methods for handling harsh conditions such as impulse noise and coherent signals. Summary of the Invention

[0005] The technical problem to be solved by this invention is:

[0006] Current two-dimensional DOA direction finding methods mostly focus on subspace decomposition or constructing propagation operators, which cannot effectively handle the direction finding problems of impulse noise and coherent signals in harsh electromagnetic environments.

[0007] Therefore, this invention provides a two-dimensional coherent DOA estimation method under strong impulse noise environment.

[0008] The technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows:

[0009] This invention differs from common two-dimensional direction-finding methods that employ subspace decomposition and propagation operator construction. Instead, it utilizes snapshot data received from a planar array, preprocessed with infinity norm normalization, to construct a Toeplitz matrix for decoherence. Simultaneously, it introduces the Cauchy kernel correlation entropy method to further suppress the influence of impulse noise, combining this with a weighted subspace fitting method for direction finding. Furthermore, the proposed quantum swallow optimization mechanism combines continuous quantum optimization with the swallow algorithm, improving the convergence speed of the swallow algorithm while optimizing its breeding and death processes. This allows the quantum swallow optimization mechanism to solve the objective function designed by the weighted subspace fitting method more quickly and accurately. This invention not only achieves two-dimensional DOA estimation under impulse noise conditions but also processes coherent signals, improving direction-finding accuracy. It also optimizes the search capability and convergence speed of the swallow algorithm, enabling rapid and accurate two-dimensional coherent signal direction finding under impulse noise conditions.

[0010] This invention provides a method for estimating the direction of arrival (DOA) of a two-dimensional coherent signal source under strong impulse noise conditions, comprising the following steps:

[0011] Step 1: Establish a two-dimensional sampling signal model of a uniform area array under impulse noise environment, and obtain snapshot data of the area array;

[0012] Step 2: Construct a covariance matrix based on the infinite norm Cauchy kernel correlation entropy using the snapshot data received by the uniform area array, and obtain the weighted subspace fitting equation based on the infinite norm Cauchy kernel correlation entropy.

[0013] Step 3: Construct a fitness function based on the weighted subspace fitting equation based on the infinite norm Cauchy kernel correlation entropy, initialize the quantum swallow population, including their respective quantum positions, calculate the mapping positions of the quantum swallows and their corresponding fitness values, and determine the global optimal quantum position;

[0014] Step 4: Update the quantum position of each quantum swiftlet according to the migration strategy of the quantum swiftlet population and calculate the mapped position;

[0015] Step 5: Update the quantum position of each quantum swiftlet according to the breeding strategy of the quantum swiftlet population and calculate the mapping position;

[0016] Step 6: Calculate the fitness value of the quantum swallow based on the updated quantum swallow's position, and update the global optimal quantum position according to the quantum swallow's fitness value;

[0017] Step 7: Update the population size based on the mortality strategy of the quantum wild swift;

[0018] Step 8: Repeat steps 4 to 7 until the maximum number of iterations is reached, output the global optimal position, and obtain the estimated direction of the incoming wave signal through mapping.

[0019] Furthermore, step one includes the following steps:

[0020] Assume an array has the following number of elements: A uniform array, wherein the spacing between adjacent array elements is . , Let be the wavelength of the incident signal, then... A far-field narrowband signal is incident on the array, and the two-dimensional wave direction of arrival is... , ,in For the first The elevation angle of the incident signal, For the first Given the azimuth angle of an incident signal, and selecting the first element of the first row and first column of a uniform surface array as the reference element, then in Time of the first Line 1 The array element received the first The signal is ,in, For the incident signal, For the first Line 1 The impulse noise signal received by the array elements follows a SaS distribution; the array receives the first... The two-dimensional signal model of the subsample is represented as: ,in , This represents the maximum number of snapshots; in this formula... ,in Indicates the first The snapshot data received by the array elements Indicates the first Line 1 The array element received the first Second snapshot data, , ; For array guiding matrix, where For the first The steering vector corresponding to each information source For the guiding matrix about On the axis The steering vector of each information source, For the guiding matrix about On the axis The steering vector of each information source, It is the elevation angle vector of the incoming wave signal. It is the azimuth vector of the incoming wave signal. Indicates the Kronecker product; For signal vectors, Let Sa be the impulse noise vector that follows a SaS distribution, where , Indicates the first The second quick shot Line 1 The impulse noise received by the array elements of the column.

[0021] Furthermore, step two includes the following steps:

[0022] Using the snapshot data received from the entire area array, a Toeplitz matrix based on the infinite-norm Cauchy kernel correlation entropy is constructed as the covariance matrix. ,in Block matrix middle and The calculation method is as follows ,in , express conjugate, This represents the absolute value operation. Represents the Cauchy kernel function. ,in This represents the kernel length of the Cauchy kernel function. express The weighting constant between them express The conjugate transpose of the matrix; for the covariance matrix Eigenvalue decomposition yields ,in It is the signal subspace spanned by the vectors corresponding to the large eigenvalues. It is the noise subspace spanned by the vectors corresponding to small eigenvalues. These are the characteristic values ​​corresponding to the signal portion. These are the eigenvalues ​​corresponding to the noise component, and the orthogonal projection matrix is ​​constructed as follows. ,in Given the array steering matrix, the angle estimation equation constructed using the weighted subspace fitting method is: ,in ,in It is the average of the small eigenvalues ​​corresponding to the noise component. It is the identity matrix. A function for finding the trace of a matrix.

[0023] Furthermore, step three includes the following steps:

[0024] Assuming it exists initially Only quantum wild swallows, the first The size of the quantum wild swift population at the next iteration is , ,in This represents the initial population size of the quantum wild swift. The minimum population size of the quantum wild swift after mortality and elimination; define the first The first generation The quantum position of a single quantum wild swallow is ,in , , , To find the maximum dimension of the solution space, The number of incident sources is given; the mapped position of the quantum swallow is obtained through mapping. The mapping equation is , ,in For the mapping position number The lower bound of the dimension, For the mapping position number The upper limit of the dimension will be the first The first generation obtained by the sub-mapping Only the quantum wild swallow mapping position is substituted into the fitness function for fitness evaluation. The fitness equation is: ,in , Remember to optimize search to the first The globally optimal position for the entire quantum wild swift population with the best fitness value up to generation is: .

[0025] Furthermore, step four includes the following steps:

[0026] During the migration of the quantum wild swift colony, the first During the nth iteration Only the quantum wild swallow The quantum rotation angle update equation for dimension is: ,in , , and All indicate Random numbers that are uniformly distributed between them. Indicates migration factor, , , and Indicates the first During the nth iteration Only, the first Only, the first Only passed The first quantum position of the wild swallow Dimension; If the current quantum swallow is the best, second best, or last best quantum swallow after ranking, then select its corresponding first-ranked quantum swallow. Only, the first Only passed When only quantum wild swallows are used, the position is supplemented according to the head-and-tail supplementation principle; a simplified simulated quantum rotating gate is used to update the position. The generation Only quantum wild swallow quantum position, quantum position number The update equation for dimension is: ,in For the first The first generation The first quantum wild swallow quantum rotation angle The updated quantum position of the wild swift is mapped to obtain the mapped position.

[0027] Furthermore, step five includes the following steps:

[0028] In the During each iteration, each quantum swallow will produce one offspring. The quantum position of a single offspring of a quantum wild swallow is denoted as ,like Then the quantum rotation angle of the offspring quantum wild swallow The dimensional update equation is ,in Indicates the scaling factor. and All indicate Random numbers that are uniformly distributed between them. Indicates reproductive factor, express random integers between [a certain range] Indicates the first Quantum wild swift population size at the next iteration Indicates until the The first globally optimal quantum position Dimension; otherwise, the quantum rotation angle of the offspring quantum swallow is the first dimension. The dimensional update equation is ,in express Random numbers that are uniformly distributed between them. and express Random integers between; quantum position number The dimensional update equation is ;if ,but ,in express Random numbers that are uniformly distributed between them. Evolutionary factor; otherwise Finally, regarding the first The generation The quantum position of a quantum wild swallow Mapping to obtain the mapping position .

[0029] Furthermore, step seven includes the following steps:

[0030] After each iteration, the quantum wild swift will eliminate individuals with low fitness values ​​to reduce the population size. The population size of quantum wild swiftlets at the next iteration ,in, The mortality constant, This represents the rounding operation.

[0031] This invention provides a two-dimensional coherent signal source direction-of-arrival estimation system under strong impulse noise environment. The system has a program module corresponding to the steps of the method described in any of the above technical solutions, and executes the steps in the above-described two-dimensional coherent signal source direction-of-arrival estimation method under strong impulse noise environment when running.

[0032] The present invention provides a computer-readable storage medium storing a computer program configured to, when called by a processor, implement the steps in the method for estimating the direction of arrival of a two-dimensional coherent signal source under strong impulse noise environment as described in any of the above technical solutions.

[0033] Compared with the prior art, the beneficial effects of the present invention are:

[0034] (1) This invention utilizes the snapshot data received by the uniform area array to construct a covariance matrix based on the infinite norm Cauchy kernel correlation entropy and combines it with a weighted subspace fitting method to achieve decoherence while suppressing the influence of impulse noise. It successfully realizes the direction finding of two-dimensional coherent signals under strong impulse noise environment, avoids the accuracy loss caused by subspace decomposition and the construction of propagation operators, and further improves the accuracy of direction finding.

[0035] (2) The quantum swallow optimization mechanism proposed in this invention combines continuous quantum optimization theory with swallow algorithm, realizes effective solution of the objective function designed by weighted subspace fitting method, improves the breeding process and death process of swallow algorithm, increases the search capability of quantum swallow optimization mechanism, and further improves the convergence speed of algorithm by introducing simulated quantum rotation gate and quantum coding, and can realize fast and accurate solution of incoming wave signal direction, so that the wave arrival estimation result has better real-time performance. Attached Figure Description

[0036] Figure 1 This is a flowchart illustrating the two-dimensional coherent signal source direction-of-arrival estimation method under strong impulse noise environment in an embodiment of the present invention.

[0037] Figure 2This is a schematic diagram of the direction finding results of two coherent sources under weak impulse noise environment in an embodiment of the present invention;

[0038] Figure 3 This is a schematic diagram of the direction finding results of two coherent sources under a strong impulse noise environment in an embodiment of the present invention;

[0039] Figure 4 The graphs show the relationship between the root mean square error and the generalized signal-to-noise ratio for the two direction-of-arrival estimation methods in this invention embodiment.

[0040] Figure 5 The graphs show the relationship between the root mean square error and the number of snapshots for the two direction-of-arrival estimation methods in this invention. Detailed Implementation

[0041] To enable those skilled in the art to better understand the present invention, exemplary embodiments or examples of the present invention will be described below in conjunction with the accompanying drawings. Obviously, the described embodiments or examples are merely some, not all, of the embodiments or examples of the present invention. All other embodiments or examples obtained by those skilled in the art based on the embodiments or examples of the present invention without inventive effort should fall within the scope of protection of the present invention.

[0042] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0043] This invention presents a two-dimensional coherent signal estimation method under strong impulse noise conditions. This method suppresses the influence of impulse noise while achieving decoherence by constructing a correlation entropy covariance matrix based on the infinite-norm Cauchy kernel. Combined with a weighted subspace fitting method, it achieves two-dimensional DOA estimation under extreme noise conditions. A quantum swift optimization mechanism is proposed, combining the wild swallow algorithm and continuous quantum optimization mechanism. The breeding and death processes of the wild swallow algorithm are improved to make it more suitable for multi-dimensional search. Quantum encoding and simulated quantum evolution equations are used to further improve the convergence performance of the algorithm, enabling rapid and accurate two-dimensional DOA direction finding under harsh environments such as coherent signals and strong impulse noise. Experimental simulations verify that the two-dimensional DOA estimation method of this invention can accurately and effectively achieve two-dimensional DOA estimation under coherent signal and strong impulse noise conditions.

[0044] Combination Figure 1 As shown, this invention provides a method for estimating the direction of arrival (DOA) of a two-dimensional coherent signal source under strong impulse noise conditions, comprising the following steps:

[0045] Step 1: Establish a two-dimensional sampling signal model of a uniform area array under impulse noise environment, and obtain snapshot data of the area array.

[0046] Assume an array has the following number of elements: A uniform array, wherein the spacing between adjacent array elements is . , Let be the wavelength of the incident signal. A far-field narrowband signal is incident on the array, and its two-dimensional wave direction of arrival is , ,in For the first The elevation angle of the incident signal, For the first The azimuth angle of the incident signal. Taking the first row and first column of the uniform surface array as the reference element, then in... Time of the first Line 1 The array element received the first The signal is ,in, For the incident signal, For the first Line 1 The array elements receive impulse noise signals that follow a SaS distribution. The array received the first The two-dimensional signal model of the subsampled signal can be represented as: ,in , This represents the maximum number of snapshots; in this formula... ,in Indicates the first The snapshot data received by the array elements Indicates the first Line 1 The array element received the first Second snapshot data, , ; For array guiding matrix, where For the first The steering vector corresponding to each information source For the guiding matrix about On the axis The steering vector of each information source, For the guiding matrix about On the axis The steering vector of each information source, It is the elevation angle vector of the incoming wave signal. It is the azimuth vector of the incoming wave signal. Indicates the Kronecker product. ; For signal vectors, Let Sa be the impulse noise vector that follows a SaS distribution, where , Indicates the first The second quick shot Line number The impulse noise received by the array elements of the column.

[0047] Step 2: Construct a covariance matrix based on the infinite norm Cauchy kernel correlation entropy using the snapshot data received by the uniform area array, and obtain the weighted subspace fitting equation based on the infinite norm Cauchy kernel correlation entropy.

[0048] Using the snapshot data received from the entire area array, a Toeplitz matrix based on the infinite-norm Cauchy kernel correlation entropy is constructed as the covariance matrix. ,in Block matrix middle and The calculation method is as follows ,in , express conjugate, Indicates the first Line number The array element received the first Second snapshot data, This represents the absolute value operation. Represents the Cauchy kernel function. ,in This represents the kernel length of the Cauchy kernel function. express The weighting constant between them express The conjugate transpose of the matrix. , For the covariance matrix Eigenvalue decomposition can yield ,in It is the signal subspace spanned by the vectors corresponding to the large eigenvalues. It is the noise subspace spanned by the vectors corresponding to small eigenvalues. These are the characteristic values ​​corresponding to the signal portion. These are the eigenvalues ​​corresponding to the noise component. The orthogonal projection matrix is ​​constructed as follows: ,in Given the array steering matrix, the angle estimation equation constructed using the weighted subspace fitting method is: ,in ,in It is the average of the small eigenvalues ​​corresponding to the noise component. It is the identity matrix. A function for finding the trace of a matrix.

[0049] Step 3: Construct a fitness function based on the weighted subspace fitting equation based on the infinite norm Cauchy kernel correlation entropy, initialize the quantum swallow population, including its respective quantum position, calculate the mapping position of the quantum swallow and the corresponding fitness value, and determine the global optimal quantum position.

[0050] Assuming it exists initially The population size of the quantum wild swift changes continuously with the iteration process, reaching its peak at the [number]th iteration. The size of the quantum wild swift population at the next iteration is , ,in This represents the initial population size of the quantum wild swift. This represents the minimum population size of the quantum wild swift after mortality and elimination. Each quantum wild swift has its own unique quantum position, defined as the quantum position during the algorithm iteration process. The first generation The quantum position of a single quantum wild swallow is ,in , , , To find the maximum dimension of the solution space, Let be the number of incident sources. The mapped position of the quantum swallow is obtained through mapping. The mapping equation is , ,in For the mapping position number The lower bound of the dimension, For the mapping position number The upper limit of the dimension. The first... The first generation obtained by the sub-mapping Only the quantum wild swallow mapping position is substituted into the fitness function for fitness evaluation. The fitness equation is: ,in The search will be optimized up to the [number]th [item]. The position with the best fitness value in the entire quantum wild swift population up to generation is denoted as the global optimal position. .

[0051] Step 4: Update the quantum position of each quantum swiftlet according to the migration strategy of the quantum swiftlet population and calculate the mapped position.

[0052] In the In the iteration, the quantum swallows are ranked from best to worst according to their fitness values. During the migration of the quantum swallow population, the current migration direction of a quantum swallow is influenced by the positions of surrounding quantum swallows. Therefore, the... During the nth iteration Only the quantum wild swallow The quantum rotation angle update equation for dimension is: ,in , , and All indicate Random numbers that are uniformly distributed between them. Indicates migration factor, , , and Indicates the first During the nth iteration Only, the first Only, the first Only passed The first quantum position of the wild swallow dimension, , If the current quantum swallow is the best, second best, or last best quantum swallow after ranking, then select its corresponding [rank / ranking]. Only, the first Only passed When only quantum wild swallows are used, the principle of head-tail complementarity is followed, that is, the optimal quantum wild swallow's first... Only, the first The position of a single quantum swallow is supplemented using the last two quantum swallow positions; the second-best quantum swallow's position is... Only the last quantum swallow's position is used to supplement the position; the last quantum swallow's position is the... Only quantum swallows are supplemented using the optimal quantum swallow position. The simplified simulated quantum rotating gate is used to update the... The generation Only quantum wild swallow quantum position, quantum position number The update equation for dimension is: ,in For the first The first generation The first quantum wild swallow quantum rotation angle Dimension. Finally, through the analysis of the first... The generation The quantum position of a quantum wild swallow Mapping to obtain the mapping position Mapping equation ,in For the mapping position number The lower bound of the dimension, For the mapping position number The upper limit of the dimension, , .

[0053] Step 5: Update the quantum position of each quantum swiftlet according to the breeding strategy of the quantum swiftlet population and calculate the mapping position.

[0054] The quantum position of the quantum swiftlet was updated again based on its breeding strategy. During each iteration, each quantum swallow will produce one offspring. The quantum position of a single offspring of a quantum wild swallow is denoted as .if Then the quantum rotation angle of the offspring quantum wild swallow The dimensional update equation is ,in Indicates the scaling factor. and All indicate Random numbers that are uniformly distributed between them. Indicates reproductive factor, express random integers between [a certain range] Indicates the first The quantum wild swift population size at the next iteration. Indicates until the The first globally optimal quantum position dimension, , Otherwise, the quantum rotation angle of the offspring quantum wild swallow... The dimensional update equation is ,in express Random numbers that are uniformly distributed between them. and express A random integer between [values]. Update the [value] using a simplified simulated quantum rotation gate. The generation Only the offspring quantum wild swallow quantum position, quantum position number The dimensional update equation is ,in For the first The first generation The first quantum wild swallow quantum rotation angle Wei. If Then the quantum position of the offspring quantum swallow is used instead of the first. The quantum position of a quantum wild swallow, that is ,in express Random numbers that are uniformly distributed between them. Evolutionary factor; otherwise Finally, regarding the first... The generation The quantum position of a quantum wild swallow Mapping to obtain the mapping position Mapping equation ,in For the mapping position number The lower bound of the dimension, For the mapping position number The upper limit of the dimension, , .

[0055] Step Six: Place the first The generation The mapping position of only quantum wild swallows Substitute into fitness function Calculate the corresponding fitness value, where The global optimal quantum position is updated based on the fitness value of the quantum swallow. .

[0056] Step 7: Update the population size based on the mortality strategy of the quantum wild swift.

[0057] During migration, the population size of Quantum Swifts gradually decreases. After each iteration, Quantum Swifts eliminate individuals with low fitness values ​​to reduce the population size. The population size of quantum wild swiftlets at the next iteration ,in This represents the initial population size of the quantum wild swift. This represents the minimum population size of the quantum wild swift. The mortality constant, This represents the rounding operation.

[0058] Step 8: Determine if the maximum number of iterations has been reached. If the target position is not reached, return to step four; otherwise, proceed to the next step; output the globally optimal position. The mapped position is the estimated direction of the incoming wave signal.

[0059] The proposed method (algorithm) for estimating the direction of arrival of a two-dimensional coherent signal source under strong impulse noise environment is the underlying technical core of this invention, and various products can be derived based on the algorithm.

[0060] Based on the method proposed in this invention, a two-dimensional coherent signal source direction-of-arrival estimation system under strong impulse noise environment is developed using a programming language. The system has program modules corresponding to the steps of the above technical solution, and executes the steps in the above-described method for estimating the direction of arrival of a two-dimensional coherent signal source under strong impulse noise environment when running.

[0061] The developed system (software) computer program is stored on a computer-readable storage medium. This computer program is configured to implement the steps of the above-described method for estimating the direction of arrival (DOA) of a two-dimensional coherent signal source under strong impulse noise conditions when called by a processor. In other words, the invention is materialized on a carrier, becoming a computer program product.

[0062] Various implementations of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, application-specific integrated circuits (ASICs), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include: implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0063] The computational programs (also referred to as programs, software, software applications, or code) of this invention include machine instructions of a programmable processor and can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, device, and / or apparatus (e.g., disk, optical disk, memory, programmable logic device PLD) for providing machine instructions and / or data to a programmable processor, including machine-readable media that receive machine instructions as machine-readable signals. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0064] The beneficial effects of the present invention will be described below with reference to specific embodiments.

[0065] Example 1

[0066] In this embodiment, the two-dimensional coherent signal estimation method under strong impulse noise environment proposed in this invention is hereinafter referred to as QWG-INCK-WSF. A two-dimensional coherent signal direction finding method under strong impulse noise environment is simulated. First, the parameters of the uniform area array system are set, including the number of rows of the area array. , number of columns Array element spacing Set the number of information sources. The directions of incoming waves are respectively and Cauchy kernel length Weighting constant The initial population size of the quantum wild swift Minimum population size Maximum number of iterations ,forward The upper and lower bounds of the dimensional mapping position are ,back The upper and lower bounds of the dimensional mapping position are Migration factors scaling factor Reproductive factors Evolutionary factors Death constant .

[0067] To further verify the superiority of the direction-finding method of this invention, this embodiment compares it with the algorithm of this invention using a two-dimensional multiple signal classification algorithm based on fractional low-order covariance moments, abbreviated as FLOC-2D-MUISC. This algorithm references existing models. [1] .

[0068] Figure 2 This diagram shows the direction finding results under the condition of two coherent signal sources, with a generalized signal-to-noise ratio of... Quick shot number Noise environment is With weak impulse noise, the order of the fractional low-order covariance moment is 0.9, and a total of 30 Monte Carlo experiments were conducted. From the simulation... Figure 2 As can be seen, the direction finding results of the FLOC-2D-MUISC method are scattered due to the influence of coherent signals, while the QWG-INCK-WSF method proposed in this invention can better achieve two-dimensional DOA estimation of two coherent sources, and the direction finding results are concentrated near the true direction of arrival.

[0069] Figure 3 This diagram shows the direction finding results under the condition of two coherent signal sources, with a generalized signal-to-noise ratio of... Quick shot number Noise environment is The simulation results showed strong impulse noise, with the fractional low-order covariance moment having an order of 0.6. A total of 30 Monte Carlo experiments were conducted. Figure 3 The results show that due to the combined influence of strong impulse noise and coherent sources, the direction finding results of the FLOC-2D-MUISC method are scattered in various places, with only some direction finding results located near the true direction of arrival. In contrast, the QWG-INCK-WSF method proposed in this invention can better achieve DOA estimation of two coherent sources under strong impulse noise conditions, and the direction finding results are mostly concentrated near the true source.

[0070] Figure 4 This indicates that when two coherent signal sources are incident, the QWG-INCK-WSF method and the FLOC-2D-MUISC method of this invention are used simultaneously for processing, and the impact of the generalized signal-to-noise ratio on the direction-finding performance is compared. (Snapshot count) Noise environment is The strong impulse noise has a fractional low-order covariance moment order of 0.6. Thirty Monte Carlo experiments were performed at each generalized signal-to-noise ratio. From the simulation... Figure 4 The results show that the root mean square error of the FLOC-2D-MUISC method is improved with the increase of the generalized signal-to-noise ratio, but the overall root mean square error is still high. The root mean square error of the QWG-INCK-WSF method proposed in this invention decreases rapidly with the increase of the generalized signal-to-noise ratio, indicating that the method proposed in this invention has better estimation performance than the FLOC-2D-MUISC algorithm.

[0071] Figure 5 This indicates that, under the condition of two coherent signal sources, the QWG-INCK-WSF method and the FLOC-2D-MUISC method of this invention are used simultaneously for processing, and the impact of the number of snapshots on the direction finding performance is compared. The generalized signal-to-noise ratio is... Noise environment is The strong impulse noise results in a fractional low-order covariance moment of order 0.6. Thirty Monte Carlo experiments are performed per snapshot. From the simulation... Figure 5 The results show that the root mean square error of the FLOC-2D-MUISC method improves with the increase of the number of snapshots, but the overall root mean square error remains at a high level. The root mean square error of the QWG-INCK-WSF method proposed in this invention does not change much with the increase of the number of snapshots, and it still has a lower root mean square error even under low snapshot conditions. This indicates that the method proposed in this invention can still obtain accurate direction finding results under low snapshot conditions.

[0072] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.

[0073] [1]Zha D and Qiu T. Underwater sources location in non-Gaussian impulsivenoise environments. "Digital Signal Processing". 2006, 16(2):149-163.

Claims

1. A method for estimating the direction of arrival (DOA) of a two-dimensional coherent signal source under strong impulse noise conditions, characterized in that, Includes the following steps: Step 1: Establish a two-dimensional sampling signal model of a uniform area array under impulse noise environment, and obtain snapshot data of the area array; Step 2: Construct a covariance matrix based on the infinite norm Cauchy kernel correlation entropy using the snapshot data received by the uniform area array, and obtain the weighted subspace fitting equation based on the infinite norm Cauchy kernel correlation entropy. Step 3: Construct a fitness function based on the weighted subspace fitting equation based on the infinite norm Cauchy kernel correlation entropy, initialize the quantum swallow population, including their respective quantum positions, calculate the mapping positions of the quantum swallows and their corresponding fitness values, and determine the global optimal quantum position; Step 4: Update the quantum position of each quantum swiftlet according to the migration strategy of the quantum swiftlet population and calculate the mapped position; Step 5: Update the quantum position of each quantum swiftlet according to the breeding strategy of the quantum swiftlet population and calculate the mapping position; Step 6: Calculate the fitness value of the quantum swallow based on the updated quantum swallow's position, and update the global optimal quantum position according to the quantum swallow's fitness value; Step 7: Update the population size based on the mortality strategy of the quantum wild swift; Step 8: Repeat steps 4 to 7 until the maximum number of iterations is reached, output the global optimal position, and obtain the estimated direction of the incoming wave signal through mapping.

2. The method according to claim 1, characterized in that, Step one includes the following steps: Assume an array has the following number of elements: A uniform array, wherein the spacing between adjacent array elements is . , Let be the wavelength of the incident signal, then... A far-field narrowband signal is incident on the array, and the two-dimensional wave direction of arrival is... , ,in For the first The elevation angle of the incident signal, For the first Given the azimuth angle of an incident signal, and selecting the first element of the first row and first column of a uniform surface array as the reference element, then in Time of the first Line number The array element received the first The signal is ,in, For the incident signal, For the first Line number The impulse noise signal received by the array elements follows a SaS distribution; the array receives the first... The two-dimensional signal model of the subsample is represented as: ,in , This represents the maximum number of snapshots; in this formula... ,in Indicates the first The snapshot data received by the array element Indicates the first Line number The array element received the first Second snapshot data, , ; Let be the array guiding matrix, where For the first The steering vector corresponding to each information source For the guiding matrix about On the axis The steering vector of each information source, For the guiding matrix about On the axis The steering vector of each information source, It is the elevation angle vector of the incoming wave signal. It is the azimuth vector of the incoming wave signal. Indicates the Kronecker product; For signal vectors, Let Sa be the impulse noise vector that follows a SaS distribution, where , Indicates the first The second quick shot Line number The impulse noise received by the array elements of the column.

3. The method according to claim 2, characterized in that, Step two includes the following steps: Using the snapshot data received from the entire area array, a Toeplitz matrix based on the infinite-norm Cauchy kernel correlation entropy is constructed as the covariance matrix. ,in Block matrix middle and The calculation method is as follows ,in , express conjugate, This represents the absolute value operation. Represents the Cauchy kernel function. ,in This represents the kernel length of the Cauchy kernel function. express The weighting constant between them express The conjugate transpose of the matrix; for the covariance matrix Eigenvalue decomposition yields ,in It is the signal subspace spanned by the vectors corresponding to the large eigenvalues. It is the noise subspace spanned by the vectors corresponding to small eigenvalues. These are the characteristic values ​​corresponding to the signal portion. These are the eigenvalues ​​corresponding to the noise component, and the orthogonal projection matrix is ​​constructed as follows. ,in Given the array steering matrix, the angle estimation equation constructed using the weighted subspace fitting method is: ,in ,in It is the average of the small eigenvalues ​​corresponding to the noise component. It is the identity matrix. A function for finding the trace of a matrix.

4. The method according to claim 3, characterized in that, Step three includes the following steps: Assuming it exists initially Only quantum wild swallows, the first The size of the quantum wild swift population at the next iteration is , ,in This represents the initial population size of the quantum wild swift. The minimum population size of the quantum wild swift after mortality and elimination; define the first The first generation The quantum position of a single quantum wild swallow is ,in , , , To find the maximum dimension of the solution space, The number of incident sources is given; the mapped position of the quantum swallow is obtained through mapping. The mapping equation is , ,in For the mapping position number The lower bound of the dimension, For the mapping position number The upper limit of the dimension will be the first The first generation obtained by the sub-mapping Only the quantum wild swallow mapping position is substituted into the fitness function for fitness evaluation. The fitness equation is: ,in , Remember to optimize search to the first The globally optimal position for the entire quantum wild swift population with the best fitness value up to generation is: .

5. The method according to claim 4, characterized in that, Step four includes the following steps: During the migration of the quantum wild swift colony, the first During the nth iteration Only the quantum wild swallow The quantum rotation angle update equation for dimension is: ,in , , and All indicate Random numbers that are uniformly distributed between them. Indicates migration factor, , , and Indicates the first During the nth iteration Only, the first Only, the first Only passed The first quantum position of the wild swallow Dimension; If the current quantum swallow is the best, second best, or last best quantum swallow after ranking, then select its corresponding rank. Only, the first Only passed When only quantum wild swallows are used, the position is supplemented according to the head-and-tail supplementation principle; a simplified simulated quantum rotating gate is used to update the position. The generation Only quantum wild swallow quantum position, quantum position number The update equation for dimension is: ,in For the first The first generation The first quantum wild swallow quantum rotation angle The updated quantum position of the wild swift is mapped to obtain the mapped position.

6. The method according to claim 5, characterized in that, Step five includes the following steps: In the During each iteration, each quantum swallow will produce one offspring. The quantum position of a single offspring of a quantum wild swallow is denoted as ,like Then the quantum rotation angle of the offspring quantum wild swallow The dimensional update equation is ,in Indicates the scaling factor. and All indicate Random numbers that are uniformly distributed between them. Indicates reproductive factor, express random integers between [a certain range] Indicates the first The quantum wild swift population size at the next iteration. Indicates until the The first globally optimal quantum position Dimension; otherwise, the quantum rotation angle of the offspring quantum swallow is the first dimension. The dimensional update equation is ,in express Random numbers that are uniformly distributed between them. and express Random integers between; quantum position number The dimensional update equation is ;if ,but ,in express Random numbers that are uniformly distributed between them. Evolutionary factor; otherwise Finally, regarding the first The generation The quantum position of a quantum wild swallow Mapping to obtain the mapping position .

7. The method according to claim 6, characterized in that, Step seven includes the following steps: After each iteration, the quantum wild swift will eliminate individuals with low fitness values ​​to reduce the population size. The population size of quantum wild swiftlets at the next iteration ,in, The mortality constant, This represents the rounding operation.

8. A two-dimensional coherent signal source direction-of-arrival estimation system under strong impulse noise environment, characterized in that, The system has a program module corresponding to the steps of the method described in any one of claims 1 to 7, and executes the steps in the above-described method for estimating the direction of arrival of a two-dimensional coherent signal source under strong impulse noise environment when it is run.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program configured to, when invoked by a processor, implement the steps in the method for estimating the direction of arrival of a two-dimensional coherent signal source under strong impulse noise environment as described in any one of claims 1 to 7.

Images