Underdetermined multipath arrival angle estimation method based on high-order statistics and non-uniform array
Through the multipath arrival angle estimation method based on high-order statistics and non-uniform arrays, the problems of low estimation accuracy and high computational complexity in low signal-to-noise ratio environments are solved, and high-precision multipath arrival angle in underdetermined scenarios are achieved Estimation, optimized array design with hybrid integer programming improves estimation performance and flexibility.
Patent Information
- Application Number
- CN202210470513.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-04-28
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2042-04-28
AI Technical Summary
The prior art has low estimation accuracy in low signal-to-noise ratio environments, and cannot effectively estimate the arrival angle in underdetermined scenarios, and has high computational complexity, so it is impossible to achieve a larger array diameter to improve accuracy.
Using multipath arrival angle estimation method based on high-order statistics and non-uniform arrays, by acquiring channel data, calculating the spatial high-order accumulation amount, constructing a covariance matrix, performing singular value decomposition or feature decomposition, defining a virtual array manifold, and combining signals The orthogonality of subspace and noise subspace searches spectral peaks, and optimizes array design using mixed integer programming.
It realizes high-precision multipath arrival angle estimation in low signal-to-noise ratio environments, reduces the calculation complexity, and can accurately estimate the number of multipaths in underdetermined scenarios, which has strong robustness and flexibility.
Smart Images

Figure CN114879134B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of large-scale multi-input multi-output array signal processing, and more specifically, relates to an underdetermined multipath arrival angle estimation method based on high-order statistics and non-uniform arrays. Background Art
[0002] With the advent of the 5G era, intelligent applications are becoming more and more widespread, and more and more fields require automatic direction positioning. Therefore, the estimation of the direction of arrival has received great attention. The accuracy of the estimation of the direction of arrival directly affects the performance of upper-layer applications. At the same time, the spatial signal propagation environment is becoming more and more complex, and the number of directions of arrival is often greater than the number of elements in the antenna array. How to achieve the estimation of the direction of arrival in the underdetermined scenario is a key issue to reduce the complexity of the antenna system and improve the system performance.
[0003] Common methods for arrival angle estimation are subspace methods, which are mainly divided into algorithms based on the orthogonality of signal subspace and noise subspace and algorithms based on the rotational invariance of signal subspace. Both methods are based on second-order statistics to calculate the mixed covariance matrix of far-field uncorrelated mixed source signals. However, the reception and transmission of signals in space will introduce Gaussian noise, and second-order statistics do not have the ability to suppress noise. This leads to a significant decrease in the accuracy of these two methods in scenarios with low signal-to-noise ratio. At the same time, based on the Nyquist sampling theorem and the stationary random signal assumption, the uniform linear array (ULA) is the most commonly used antenna array structure in the arrival angle estimation problem. It is well known that the larger the antenna aperture, the higher the accuracy of the arrival angle estimation. However, if a uniform linear array wants to obtain a large array aperture, a large number of array elements are required, which in turn causes a high system overhead.
[0004] In addition, the arrival angle estimation method based on second-order statistics and ULA can only estimate the number of incoming waves that is less than the number of ULA array elements. The actual situation is complicated, and the number of incoming wave directions is often greater than the number of antenna array elements, which is underdetermined. In this case, the ULA array cannot estimate the exact number of incoming wave directions. Summary of the invention
[0005] In view of the defects of the prior art and the need for improvement, the present invention provides an underdetermined multipath arrival angle estimation method based on high-order statistics and non-uniform arrays, aiming to solve the technical problems existing in the prior art such as high computational complexity, low estimation accuracy in low signal-to-noise ratio environments, and inability to effectively estimate the arrival angle in underdetermined scenarios.
[0006] To achieve the above object, in a first aspect, the present invention provides an underdetermined multipath arrival angle estimation method based on high-order statistics and non-uniform arrays, comprising:
[0007] Get channel data;
[0008] Calculating spatial high-order cumulants of the channel data and organizing them into a covariance matrix;
[0009] Constructing a corresponding virtual array manifold according to the organization mode of the covariance matrix;
[0010] Performing singular value decomposition or eigendecomposition on the covariance matrix to obtain a signal subspace and a noise subspace;
[0011] Based on the orthogonality of the signal subspace and the noise subspace, the spatial spectrum is defined in combination with the virtual array manifold, and the spectrum peak is searched to obtain the estimated value of the multipath arrival angle.
[0012] Furthermore, the non-uniform array structure is obtained by solving a mixed integer programming problem, wherein the mixed integer programming problem aims to maximize the non-repetitiveness of virtual array elements in the virtual array manifold.
[0013] Further, the high-order cumulant is a fourth-order cumulant;
[0014] The calculating of the spatial high-order cumulants of the channel data includes: calculating the spatial high-order cumulants of the channel data in a double conjugation manner.
[0015] Furthermore, the spatial spectrum Spec(θ) is expressed as:
[0016]
[0017] Where b(θ) represents the virtual array manifold, U n represents the noise subspace, b H (θ) and denote b(θ) and U respectively n is the conjugate transpose of , where θ represents the angle.
[0018] Further, performing singular value decomposition or eigendecomposition on the covariance matrix to obtain a signal subspace and a noise subspace includes:
[0019] Perform singular value decomposition or eigendecomposition on the covariance matrix, arrange the singular values or eigenvalues from large to small, select the singular vectors or eigenvectors corresponding to the first G larger singular values or eigenvalues to form a signal subspace, and the singular vectors and eigenvectors corresponding to the remaining singular values or eigenvalues to form a noise subspace; G is the number of multipaths.
[0020] Further, the searching spectrum peak to obtain an estimated value of the multipath arrival angle includes:
[0021] Determine the search range and search step size of angle θ;
[0022] For each search angle, calculate the value Spec(θ) corresponding to the spatial spectrum;
[0023] In the spatial spectrum Spec(θ), search for the corresponding number of spectral peaks from high to low, extract the angle values corresponding to the spectral peaks, and use them as the estimated values of the corresponding arrival angles.
[0024] In a second aspect, the present invention provides an underdetermined multipath arrival angle estimation device based on high-order statistics and non-uniform arrays, comprising:
[0025] An acquisition module, used for acquiring channel data;
[0026] A calculation module, used for calculating the spatial high-order cumulants of the channel data and organizing them into a covariance matrix;
[0027] A construction module, used for constructing a corresponding virtual array manifold according to the organization mode of the covariance matrix;
[0028] A decomposition module, used for performing singular value decomposition or eigendecomposition on the covariance matrix to obtain a signal subspace and a noise subspace;
[0029] The estimation module is used to define a spatial spectrum based on the orthogonality of the signal subspace and the noise subspace in combination with the virtual array manifold, and to search for spectrum peaks to obtain an estimated value of the multipath arrival angle.
[0030] In a third aspect, the present invention provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is executed by a processor, the device where the storage medium is located is controlled to execute the arrival angle estimation method as described in the first aspect.
[0031] In general, the above technical solutions conceived by the present invention can achieve the following beneficial effects:
[0032] (1) The present invention uses a non-uniform array to achieve maximum non-repetition of virtual extended array elements obtained by high-order cumulants. Unlike the method used in the present invention, existing arrival angle estimation methods either cannot achieve array expansion with a large degree of freedom under a limited number of array elements, or cannot achieve high-precision arrival angle estimation in underdetermined scenarios. At the same time, the present method also uses the excellent characteristic of high-order cumulants that theoretically automatically suppresses Gaussian noise, achieves strong robustness to noise, and can still show good estimation performance under low signal-to-noise ratio.
[0033] (2) The present invention designs the non-uniform array as a solvable mixed integer programming problem, which can be solved quickly.
[0034] (3) Compared with other arrival angle estimation methods, the design method of the present invention can be used in underdetermined scenarios and has the advantages of high estimation accuracy and low computational complexity; it has certain feasibility and practical promotion value. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] Figure 1 It is a flow chart of an underdetermined multipath arrival angle estimation method based on high-order statistics and non-uniform array provided by the present invention;
[0036] Figure 2 Schematic diagram of a non-uniform array design provided in an embodiment of the present invention; wherein (a) is a six-element non-uniform linear array structure, and (b) is a six-element non-uniform planar array structure;
[0037] Figure 3 The figure is a comparison diagram of the spatial spectrum and arrival angle estimation matching obtained by using the NULA structure proposed in the present invention and the results under the ULA structure; wherein (a) and (b) are the spatial spectrum and arrival angle estimation matching obtained by using the NULA structure proposed in the present invention, and (c) and (d) are the spatial spectrum and arrival angle estimation matching obtained under the ULA structure, respectively;
[0038] FIG. 4( a ) and FIG. 4( b ) are curve diagrams showing the relationship between the arrival angle estimation success rate and the root mean square error and the signal-to-noise ratio provided in the embodiments of the present invention. DETAILED DESCRIPTION
[0039] In order to make the purpose, system composition, technical scheme and advantages of the present invention clearer, the present invention 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 invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.
[0040] In the present invention, the terms "first", "second", etc. (if any) in the present invention and the accompanying drawings are used to distinguish similar objects but not necessarily to describe a specific order or sequence.
[0041] See also Figure 1 The present invention provides an underdetermined multipath arrival angle estimation method based on high-order statistics and non-uniform arrays. The method includes operations S1 to S4.
[0042] It should be noted that, in this embodiment, high-order statistics are third-order and higher-order statistics, and the following takes fourth-order statistics as an example for a specific description.
[0043] Operation S1: acquiring channel data.
[0044] In this embodiment, in the signal processing stage, the non-uniform array at the receiving end receives the pilot signal transmitted by the transmitting end, and obtains channel data using basic methods such as least squares.
[0045] The principle of the non-uniform array design at the signal receiving end is to maximize the non-repetitiveness of the virtual extended array elements obtained by the high-order cumulants. Specifically, the non-uniform antenna array structure is obtained by solving a mixed integer programming problem, which can be described as follows after appropriate relaxation:
[0046]
[0047]
[0048] subject to 1-INF·z m ≤q (m) D T ≤-1+INF·(1-z m )m=1,…,M-1,
[0049]
[0050] in represents the coordinates of the antenna array elements, and M is the number of antenna array elements. The diagonal elements of the (M-1)×M matrix Q are all -1, and the upper sub-diagonal elements are all 1, that is, Q i,i =-1, Q i,i+1 =1,i=1,…,M-1,q (m) Denotes the mth row of the matrix Q. U = M(M-1) / 2. J i =[0 M-i,i-1 1 M-i I M-i ],i=1,…,M-1,1 U-1 represents a (U-1)×1 column vector whose elements are all 1, j (u) represents the u-th row of matrix J, J [u] It means the matrix J after deleting the uth row, J A =[1 U-1 j (1) … 1 U-1 j (U) ] T and J B =[J [1] … J [U] ] T , [·] T represents the matrix transpose operation. INF is a reasonably large number. This problem can be quickly solved using MATLAB's CVX toolbox.
[0051] The signal receiving end processes the received signal to obtain the channel impulse response h(t,τ) or the channel transfer function H(t,f). The channel impulse response h(t,τ) can be obtained by the inverse Fourier transform of the channel transfer function H(t,f) with respect to the frequency f, that is,
[0052] Operation S2: Calculate the spatial high-order cumulants of the channel data and organize them into a covariance matrix; and construct a corresponding virtual array manifold according to the organization method of the covariance matrix.
[0053] The spatial fourth-order cumulants are calculated for the channel impulse responses h(t,τ) of different antennas. This method uses a double conjugate method to calculate the fourth-order cumulants c 4 ,Right now
[0054] There are many ways to organize the fourth-order cumulant covariance matrix. In this embodiment, the element c is placed in the (i-1)M+kth row and (j-1)M+lth column of the matrix. 4 (h i ,h j ,h k ,h l ) to organize the fourth-order cumulants into a covariance matrix R 4 , and obtain the corresponding virtual array manifold where a(θ) is the antenna array manifold, and for ULA it is d is the array element gap, λ is the wavelength of the signal. The covariance matrix R 4 The sample estimate of
[0055]
[0056] Where N is the number of samples, (·) H represents the conjugate transpose operation of the matrix, h(n) represents the nth sample of the channel data vector of M array elements, represents the Kronecker product.
[0057] In operation S3, singular value decomposition or eigendecomposition is performed on the covariance matrix to obtain a signal subspace and a noise subspace.
[0058] In this embodiment, singular value decomposition is performed on the quasi-covariance matrix As an example, the singular values are arranged from large to small, and the singular vectors corresponding to the larger singular values corresponding to the size of the signal subspace are selected to form the signal subspace U s , and obtain the noise subspace U n Specifically, the singular vectors corresponding to the first G larger singular values are selected to form the signal subspace, and the singular vectors corresponding to the remaining singular values are selected to form the noise subspace; G is the number of multipaths.
[0059] The number of multipaths can be obtained using methods such as AIC / MDL.
[0060] Operation S4: based on the orthogonality of the signal subspace and the noise subspace, a spatial spectrum is defined in combination with the virtual array manifold, and a spectrum peak is searched to obtain an estimated value of the multipath arrival angle.
[0061] Based on the orthogonal nature of the signal subspace and the noise subspace, the spatial spectrum is defined
[0062] Determine the search range for angle θ as [θ MIN ,θ MAX ], determine the search step size of angle θ as Δ. The setting of the angle search step size affects the accuracy of arrival angle estimation and the computational complexity of the method. For each search angle, calculate the value Spec(θ) corresponding to the spatial spectrum. In the spatial spectrum Spec(θ), search for the corresponding number of spectral peaks from high to low, extract the angle value corresponding to the spectral peak, and you can get the estimated value of the corresponding arrival angle.
[0063] Example:
[0064] Figure 2 The figure shows a schematic diagram of a non-uniform array design according to an embodiment of the present invention. The multipath arrival angle estimation method based on high-order statistics and non-uniform arrays of the present invention is applied to the antenna array and large-scale multi-input multi-output wireless multipath channels to improve the accuracy of multipath arrival angle estimation. In this embodiment, a six-element NULA is used as an example to verify the high-precision array direction finding capability of the high-order cumulant combined with the non-uniform array. The structure of the six-element NULA used is d=[0 2d 8d 13d 17d20d] T , where d = λ / 2, that is, the unit of the antenna element gap is set to half the wavelength of the incoming signal. Specifically, the following steps are included:
[0065] (1) In the channel data generation stage, the signal transmitter sends a pilot signal; the signal receiver receives the incoming signal and processes it to obtain the channel transfer function, and obtains the channel impulse response through a 3360-point inverse Fourier transform
[0066] (2) In the arrival angle estimation stage, the fourth-order cumulants of the channel impulse response are calculated and organized into a 36×36 fourth-order cumulant quasi-covariance matrix in the manner described above: Get the corresponding virtual array manifold
[0067] (3) Singular value decomposition, obtain the signal subspace U s , and obtain the noise subspace U n ;
[0068] (4) Defining the spatial spectrum
[0069] (5) Determine the search range of angle θ as [θ MIN ,θ MAX ], the search step is Δ. For each search angle, calculate the value Spec(θ) corresponding to the spatial spectrum. Search for the corresponding number of spectral peaks in the spatial spectrum from high to low, extract the angle value corresponding to the spectral peak, and you can get the estimated value of the corresponding arrival angle.
[0070] The present invention uses a non-uniform array to achieve maximum non-repetition of virtual array element expansion obtained by high-order cumulants. Unlike the method used in the present invention, existing arrival angle estimation methods either cannot achieve array expansion with a large degree of freedom under a limited number of array elements, or cannot achieve high-precision arrival angle estimation in underdetermined scenarios. At the same time, the present method also uses the excellent characteristic of high-order cumulants that theoretically automatically suppresses Gaussian noise, achieves strong robustness to noise, and can still show good estimation performance under low signal-to-noise ratio.
[0071] Figure 3 The figure is a comparison of the spatial spectrum and arrival angle estimation matching obtained by using the NULA structure proposed by the present invention and the results under the ULA structure. Figure 3 As shown in the figure, when ULA faces multipath wave directions whose number is far greater than the number of its array elements, it cannot completely and accurately estimate the directions of all waves, but can only estimate a small number of angles. And the estimated angles will also be affected by the interference of other waves and become biased. The spatial spectrum peak corresponding to the ULA structure also appears rounded. On the contrary, NULA can completely and accurately estimate all wave directions, and the spatial spectrum peak is also significantly sharp.
[0072] FIG4(a) and FIG4(b) are curve diagrams showing the relationship between the arrival angle estimation success rate and the root mean square error and the signal-to-noise ratio provided by the embodiments of the present invention. As shown in FIG4(a) and FIG4(b), under 20 paths, the non-uniform array structure is completely superior to the commonly used method of the ULA structure in terms of both the root mean square error and the success rate. When the SNR is higher than 5dB, a 95% estimation success rate is maintained. When the signal-to-noise ratio is lower than 0dB, the method of the present invention also maintains a success rate of more than 80%.
[0073] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A method for underdetermined multipath arrival angle estimation based on high-order statistics and non-uniform arrays. It is characterized in that include: Get channel data; Calculating spatial high-order cumulants of the channel data and organizing them into a covariance matrix; Constructing a corresponding virtual array manifold according to the organization mode of the covariance matrix; Performing singular value decomposition or eigendecomposition on the covariance matrix to obtain a signal subspace and a noise subspace; Based on the orthogonality of the signal subspace and the noise subspace, the spatial spectrum is defined in combination with the virtual array manifold, and the spectrum peak is searched to obtain an estimated value of the multipath arrival angle; The non-uniform array structure is obtained by solving a mixed integer programming problem, wherein the mixed integer programming problem aims to maximize the non-repetitiveness of virtual array elements in the virtual array manifold. The mixed integer programming problem can be described as follows after appropriate relaxation: subject to 1-INF·z m ≤q (m) D T ≤-1+INF·(1-z m )m=1,…,M-1, in, represents the coordinates of the antenna array elements, M is the number of antenna array elements, and the diagonal elements of the (M-1)×M matrix Q are all -1, and the upper sub-diagonal elements are all 1. i,i =-1, Q i,i+1 =1,i=1,…,M-1,q (m) represents the mth row of the matrix Q, U = M(M-1) / 2, J i =[0 M-i,i-1 1 M-i I M-i ],i=1,…,M-1,1 U-1 represents a (U-1)×1 column vector whose elements are all 1, j (u) represents the u-th row of matrix J, J [u] It means the matrix J after deleting the uth row, J A =[1 U-1 j (1) … 1 U-1 j (U) ] T and J B =[J [1] … J [U] ] T , [·] T Represents a matrix transpose operation.
2. The method for estimating the angle of arrival according to claim 1, It is characterized in that The high-order cumulant is a fourth-order cumulant; The calculating of the spatial high-order cumulants of the channel data includes: calculating the spatial high-order cumulants of the channel data in a double conjugation manner.
3. The method for estimating the angle of arrival according to claim 2, It is characterized in that The spatial spectrum Spec(θ) is expressed as: Where b(θ) represents the virtual array manifold, U n represents the noise subspace, b H (θ) and denote b(θ) and U respectively n is the conjugate transpose of , where θ represents the angle.
4. The method for estimating the angle of arrival according to claim 1, It is characterized in that The performing singular value decomposition or eigendecomposition on the covariance matrix to obtain a signal subspace and a noise subspace includes: Perform singular value decomposition or eigendecomposition on the covariance matrix, arrange the singular values or eigenvalues from large to small, select the singular vectors or eigenvectors corresponding to the first G larger singular values or eigenvalues to form a signal subspace, and the singular vectors or eigenvectors corresponding to the remaining singular values or eigenvalues to form a noise subspace; G is the number of multipaths.
5. The method for estimating the angle of arrival according to claim 1, It is characterized in that The searching for spectrum peaks to obtain an estimated value of the multipath arrival angle includes: Determine the search range and search step size of angle θ; For each search angle, calculate the value Spec(θ) corresponding to the spatial spectrum; In the spatial spectrum Spec(θ), search for the corresponding number of spectral peaks from high to low, extract the angle values corresponding to the spectral peaks, and use them as the estimated values of the corresponding arrival angles.
6. An underdetermined multipath arrival angle estimation device based on high-order statistics and non-uniform arrays, include: An acquisition module, used for acquiring channel data; A calculation module, used for calculating the spatial high-order cumulants of the channel data and organizing them into a covariance matrix; A construction module, used for constructing a corresponding virtual array manifold according to the organization mode of the covariance matrix; A decomposition module, used for performing singular value decomposition or eigendecomposition on the covariance matrix to obtain a signal subspace and a noise subspace; An estimation module, configured to define a spatial spectrum based on the orthogonality of the signal subspace and the noise subspace in combination with the virtual array manifold, and search for spectrum peaks to obtain an estimated value of the multipath arrival angle; The non-uniform array structure is obtained by solving a mixed integer programming problem, wherein the mixed integer programming problem aims to maximize the non-repetitiveness of virtual array elements in the virtual array manifold. The mixed integer programming problem can be described as follows after appropriate relaxation: subject to 1-INF·z m ≤q (m) D T ≤-1+INF·(1-z m )m=1,…,M-1, in, represents the coordinates of the antenna array elements, M is the number of antenna array elements, and the diagonal elements of the (M-1)×M matrix Q are all -1, and the upper sub-diagonal elements are all 1. i,i =-1, Q i,i+1 =1,i=1,…,M-1,q (m) represents the mth row of the matrix Q, U = M(M-1) / 2, J i =[0 M-i,i-1 1 M-i I M-i ],i=1,…,M-1,1 U-1 represents a (U-1)×1 column vector whose elements are all 1, j (u) represents the u-th row of matrix J, J [u] It means the matrix J after deleting the uth row, J A =[1 U-1 j (1) … 1 U-1 j (U) ] T and J B =[J [1] … J [U] ] T , [·] T Represents a matrix transpose operation.
7. A computer-readable storage medium, It is characterized in that The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed by a processor, the device where the storage medium is located is controlled to execute the arrival angle estimation method according to any one of claims 1 to 5.
