Accurate Angle of Arrival Estimation Method Based on OPTICS under Low Signal-to-Noise Ratio

Abstract

This invention discloses a method for accurate angle-of-arrival (AOA) estimation based on OPTICS under low signal-to-noise ratio (SNR) conditions. The method includes: performing a two-dimensional beam scan in the sky to explore multiple potential OAAs; creating multiple overlapping virtual subarrays and performing digital beamforming in these directions to obtain multiple virtual subarray signals; segmenting the received signal into multiple overlapping snapshots to obtain multiple possible OAA estimates; and using a point-ordering clustering structure identification method to perform cluster analysis on the estimates to obtain the final OAA estimate. The method disclosed in this invention achieves high-precision resolution of echo signals under low SNR and complex interference environments, while effectively reducing computational complexity.

Description

Technical Field

[0001] This invention relates to a method for accurate angle of arrival estimation based on OPTICS under low signal-to-noise ratio conditions, belonging to the field of signal processing technology. Background Technology

[0002] In specific data models, the MUSIC method is widely known for its efficiency in super-resolution techniques. This method operates based on the principle that the signal subspace and noise subspace are orthogonal. However, when the signal-to-noise ratio (SNR) is extremely low, the signal and noise subspaces can mix, making them difficult to distinguish. This limits the applicability of the MUSIC method to tracking mobile radio frequency interference sources in real-time systems.

[0003] To address the aforementioned issues, various approaches have been explored in the existing technologies for accurately estimating the angle of arrival (DoA) in low signal-to-noise ratio (SNR) environments. Among these, spatial spectrum estimation methods are a frequently explored direction. For example, Xu K, Xing M, Zhang R, et al. High-Accuracy DOA Estimation Algorithm at Low SNR Through Exploiting a Supervised Index [J]. IEEE Transactions on Aerospace and Electronic Systems, 2022, 58(4):3658-3665 proposed a fuzzy similarity matrix and transformation function to optimize the signal subspace, thereby improving the performance of DoA estimation; Lan X, Si W, Dong M. New Algorithm for DOAEstimation with Low SNR [C]. 2011 International Conference on Information Management, Innovation Management and Industrial... Engineering, 2011:348-351. Combining the weighted noise subspace (WNS) spatial spectrum with the inverse weighted signal subspace projection (PWSP), the spatial spectrum is reconstructed by weighting the noise and signal subspaces respectively to estimate the DoA; K. Hameed K, Tu S, Ahmed N, et al. DOA Estimation in Low SNR Environment through Coprime Antenna Arrays: An Innovative Approach by Applying Flower Pollination Algorithm[J]. Applied Sciences, 2021, 11(17):7985. The pollen transfer algorithm (FPA) is used to estimate the DoA of electromagnetic waves received in the coprime array (CSA), thereby achieving higher degrees of freedom (DOF) and resolution. However, although spatial spectrum estimation methods can theoretically significantly improve the accuracy of estimation, in practice these methods have failed to explore the possibility of improving the signal-to-noise ratio.

[0004] In addition, there are studies using neural network models for DoA estimation. For example, in Papageorgiou GK, Sellathurai M, Eldar Y C. Deep Networks for Direction-of-Arrival Estimation in Low SNR[J]. IEEE Transactions on Signal Processing, 2021, 69:3714-3729, a convolutional neural network (CNN) with two-dimensional filters was trained from multi-channel data of a real array manifold matrix under low signal-to-noise ratio conditions; Papageorgiou GK, Sellathurai M. Direction-of-Arrival Estimation in the Low-SNR Regime via a Denoising Autoencoder[C]. 2020 IEEE 21st International Workshop on Signal Processing Advances in Wireless Communications (SPAWC), 2020:1-5, a denoising autoencoder (DAE) was trained to predict DoA; Wu L, Liu ZM, Huang Z T. Deep Convolution Network for Direction of Arrival Estimation With Sparse Prior to this paper, IEEE Signal Processing Letters, 2019, 26(11):1688-1692, the DoA estimation problem was transformed into a sparse linear inverse problem, and the inverse transform was learned from a large amount of training data by training a deep convolutional network (DCN). However, neural network models are often regarded as black boxes, and it is difficult to guarantee their stable performance when the data distribution changes.

[0005] Furthermore, existing technologies employ various clustering methods to perform DoA estimation on observed samples, whether for separating multiple signal sources or improving estimation accuracy in multipath environments. However, these methods have not effectively addressed the problem of signal masking by noise under low signal-to-noise ratio conditions.

[0006] Therefore, it is necessary to conduct more in-depth research on the estimation methods of angle of arrival in low signal-to-noise ratio environments in order to obtain accurate echo signals. Summary of the Invention

[0007] To overcome the above problems, the inventors conducted in-depth research and proposed an accurate angle-of-arrival estimation method based on OPTICS under low signal-to-noise ratio conditions. This method includes the following steps:

[0008] S1. Perform a two-dimensional beam scan in the sky to explore multiple potential angles of arrival;

[0009] S2. Create multiple overlapping virtual subarrays and perform digital beamforming in these directions to obtain multiple virtual subarray signals;

[0010] S3. The received signal is divided into multiple overlapping snapshots to obtain multiple possible angle of arrival estimates. The point sorting identification clustering structure method is used to perform cluster analysis on the estimates to obtain the final angle of arrival estimate.

[0011] In a preferred embodiment, in S1, a uniform rectangular antenna array is used to scan the beam in all directions to obtain the power received in each direction, and the direction with power greater than a threshold is taken as the potential angle of arrival.

[0012] In a preferred embodiment, the power P received by the beam in a certain direction BeamScan Represented as:

[0013] P BeamScan =ω H *R xx *ω

[0014] R xx =AR S A H +σ 2 I

[0015] Where ω represents the weight vector of direction, H represents the conjugate transpose, and R... xx It is the array correlation matrix, A represents the rotation matrix, σ 2 R is the noise variance, I is the identity matrix, and R is the noise variance. s It is the autocorrelation matrix of the source signal S;

[0016] The rotation matrix A is represented as:

[0017] A=A(φ,θ)=[a(φ1,θ1),…,a(φ i ,θ i )…,a(φ P ,θ P )]

[0018] Where φ represents the azimuth angle in that direction, θ represents the elevation angle in that direction, p represents the total number of directions of the incident signal, and a(φ) i ,θ i ) indicates direction (φ) i ,θi The vector of the incident signal.

[0019] In a preferred embodiment, the threshold is set according to the following conditions:

[0020] Condition 1: Peak value within a range of -3dB from the highest power peak;

[0021] Condition 2: The maximum number of K highest peak values.

[0022] In a preferred embodiment, in S2, the virtual subarray is created by virtually dividing the uniform rectangular antenna array into multiple subarrays.

[0023] In a preferred embodiment, the size of the virtual subarray is (M-1)×(N-1).

[0024] In a preferred embodiment, in S2, for each beamforming signal, the angle of arrival is estimated using the MUSIC method, including the following sub-steps:

[0025] Obtain the autocorrelation matrix of the beamforming signal;

[0026] For the autocorrelation matrix R yy Perform eigenvalue decomposition to obtain the eigenvector matrix;

[0027] The signal is divided into two orthogonal subspaces by using eigenvectors: a signal subspace and a noise subspace. The signal of the virtual subarray is obtained based on the noise subspace using the MUSIC method.

[0028] In a preferred embodiment, in S3, each signal is divided into multiple signals to form a snapshot, the length of the divided signal is shorter than that of the original signal, and there is overlap between the divided signals.

[0029] In a preferred embodiment, all potential angles of arrival are clustered. When the signal-to-noise ratio is low, the angle of arrival estimation points form an reachability map. The horizontal axis of the reachability map is the cluster sequence, and the vertical axis is the reachability distance. Each depression in the reachability map represents a cluster. For each cluster, the point with the highest density is selected as the angle of arrival estimate for that cluster.

[0030] The beneficial effects of this invention include:

[0031] (1) Achieve high-precision solution of echo signals in low signal-to-noise ratio and complex interference environments;

[0032] (2) Effectively reduces computational complexity

[0033] (3) The entire process processes the single-captured incident signal in the digital domain, ensuring that the algorithm outperforms the traditional MUSIC algorithm in low signal-to-noise ratio environments, whether in additive white Gaussian noise (AWGN) channels or in actual multipath channels with Doppler effects. Attached Figure Description

[0034] Figure 1 The diagram illustrates a flowchart of a method for accurate angle of arrival estimation based on OPTICS under low signal-to-noise ratio according to a preferred embodiment of the present invention.

[0035] Figure 2 This diagram illustrates an reachability map in an accurate angle-of-arrival estimation method based on OPTICS under low signal-to-noise ratio conditions according to a preferred embodiment of the present invention.

[0036] Figure 3 This diagram illustrates the process of step S1 in Example 1.

[0037] Figure 4 This diagram illustrates the process of step S2 in Example 1.

[0038] Figure 5 A schematic diagram of step S3 in Example 1 is shown. Detailed Implementation

[0039] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Through these descriptions, the features and advantages of the present invention will become clearer and more apparent.

[0040] The term “exemplary” as used herein means “serving as an example, embodiment, or illustration.” Any embodiment illustrated herein as “exemplary” is not necessarily to be construed as superior to or better than other embodiments. Although various aspects of embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless specifically indicated otherwise.

[0041] This invention provides a method for accurate angle-of-arrival estimation based on OPTICS under low signal-to-noise ratio conditions, such as... Figure 1 As shown, the method includes the following steps:

[0042] S1. Perform a two-dimensional beam scan in the sky to explore multiple potential angles of arrival;

[0043] S2. Create multiple overlapping virtual subarrays and perform digital beamforming in these directions to obtain multiple virtual subarray signals;

[0044] S3. The received signal is divided into multiple overlapping snapshots to obtain multiple possible angle of arrival estimates. The point sorting identification clustering structure method is used to perform cluster analysis on the estimates to obtain the final angle of arrival estimate.

[0045] In S1, a uniform rectangular antenna array is used to scan the beam in all directions to obtain the power received in each direction, and the direction with power greater than a threshold is taken as the potential angle of arrival.

[0046] The uniform rectangular antenna array (URA) is an M×N antenna array, where M represents the number of antennas on the x-axis and N represents the number of antennas on the y-axis.

[0047] Furthermore, the power P received by the beam in a certain direction BeamScan Represented as:

[0048] P BeamScan =ω H *R xx *ω

[0049] R xx =AR S A H +σ 2 I

[0050] Where ω represents the weight vector of direction, H represents the conjugate transpose, and R... xx It is the array correlation matrix, A represents the rotation matrix, σ 2 R is the noise variance, I is the identity matrix, and R is the noise variance. s It is the autocorrelation matrix of the source signal S.

[0051] The rotation matrix A is represented as:

[0052] A=A(φ,θ)=[a(φ1,θ1),…,a(φ i ,θ i )…,a(φ P ,θ P )]

[0053] Where φ represents the azimuth angle in that direction, θ represents the elevation angle in that direction, p represents the total number of directions of the incident signal, and a(φ) i ,θ i ) indicates direction (φ) i ,θ i The vector of the incident signal is represented as

[0054]

[0055] μ i =sinφ i sinθ i v i =cosφ i sinθ i

[0056] Among them, (μ i vi ) is the stereo projection coordinate, d is the distance between any two adjacent antennas, λ is the wavelength of the processed signal, j represents the imaginary number, the superscript T indicates transpose, m and n refer to different antennas, 1≤m≤M, 1≤n≤N.

[0057] Traditionally, the direction with the highest power is considered to be the estimated angle of arrival (DoA), meaning the highest peak in the spatial spectrum is an estimate of the DoA. However, under low signal-to-noise ratio conditions, the power of the array sidelobe peaks is close to the maximum output power of the spatial spectrum. In this invention, multiple peaks exceeding a threshold are selected, rather than just the direction with the highest power, to obtain multiple potential angles of arrival.

[0058] In a preferred embodiment, the threshold is set according to the following conditions:

[0059] Condition 1: Peak value within a range of -3dB from the highest power peak;

[0060] Condition 2: The maximum number of K highest peak values.

[0061] K is chosen empirically to minimize computation.

[0062] According to the present invention, digital beamforming is performed on the signal received by each antenna to cover a 2π×π*2 region. After threshold filtering, the potential angle of arrival can be expressed as: Where, φ k "" represents the azimuth angle of the potential angle of arrival, θ k "" indicates the pitch angle of potential arrival.

[0063] In S2, the virtual subarray is created by virtually dividing the uniform rectangular antenna array into multiple subarrays.

[0064] Preferably, the size of the virtual subarray is (M-1)×(N-1), which is the size that can achieve the best estimation accuracy.

[0065] According to the present invention, each virtual subarray performs beamforming on the potential angle of arrival, and each virtual subarray generates a beamforming signal, denoted as (y1,…,y). B ), where B represents the total number of virtual subarrays.

[0066] For each beamforming signal (y1,…,y) B The MUSIC method is used to estimate the angle of arrival, resulting in multiple angle of arrival estimates. These estimates are then combined to obtain a more refined pair of angles of arrival, denoted as {φ'}. k ,θ' k},

[0067] According to the present invention, any beamforming signal y j The angle of arrival estimation using the MUSIC method includes the following sub-steps:

[0068] Obtain the autocorrelation matrix R of the beamforming signal yy , represented as:

[0069] R yy =E[y j (t)y j H (t)]

[0070] Where E[] represents the expectation, y j (t) represents the beamforming signal y j The value at time t;

[0071] For the autocorrelation matrix R yy The eigenvector matrix is ​​obtained by performing eigenvalue decomposition, and is represented as follows:

[0072] R yy =UΛU H

[0073] Where U is the eigenvector matrix and Λ is the eigenvalue matrix.

[0074] The signal is divided into two orthogonal subspaces using eigenvectors: a signal subspace and a noise subspace. Specifically, the eigenvectors are divided into P largest eigenvalues ​​and (M-1)×(N-1)-P smaller eigenvalues, and the space U corresponding to the eigenvectors of these P largest eigenvalues ​​is formed by these eigenvectors. s The space U is called the signal subspace, which is the space of eigenvectors corresponding to (M-1)×(N-1)-P smaller eigenvalues. N That is the noise subspace.

[0075] The signal from this virtual subarray is transmitted through the MUSIC spatial spectrum function P. MUSIC express:

[0076]

[0077] μ p =sinφ p sinθ p v p =cosφ p sinθ p

[0078] Wherein, the traversal range of φ and θ is (φ″±δ, θ″±δ), where δ is an empirical value, preferably 10°.

[0079] In S2, the signal-to-noise ratio (SNR) of the incident signal is enhanced by creating multiple overlapping virtual subarrays and performing digital beamforming in these directions. Furthermore, at this stage, the MUSIC method operates only within a local region of the beamformed signal, effectively reducing computational complexity.

[0080] In S3, each signal is divided into multiple signals to form a snapshot. According to the present invention, the length of the segmented signal is shorter than that of the original signal, and there is overlap between the segmented signals.

[0081] Furthermore, according to step S1, each snapshot can generate multiple potential angles of arrival (Angles of Arrival). In environments with high signal-to-noise ratios (SNR), these potential Angles of Arrival (Angles of Arrival) will cluster more closely, eventually converging at a single point, i.e., the true Angle of Arrival of the incident signal. Conversely, in environments with low SNR, these potential Angles of Arrival may be scattered over a large area, exhibiting a random distribution. To accurately estimate the actual Angle of Arrival from these potential Angles of Arrival, this invention employs a clustering method, selecting the main cluster centers as the final Angle of Arrival estimate.

[0082] Clustering not only improves the accuracy of the estimation, but also enhances its robustness under different signal-to-noise ratio conditions.

[0083] Furthermore, the Ordering Points To Identify the Clustering Structure (OPTICS) is an algorithm for discovering clustering structures in a dataset. It is particularly suitable for handling clusters with different densities and for discovering datasets with different cluster shapes and sizes. It is widely used in the field of data statistics. In this invention, its specific process will not be described in detail.

[0084] According to this invention, all potential angles of arrival are clustered, the cluster with the highest density is identified, and its center is selected as the final angle of arrival estimate. When the signal-to-noise ratio is high and the angle of arrival is easy to estimate, most angle of arrival estimation points will be concentrated on the correct angle of arrival points, resulting in high density. When the signal-to-noise ratio is low, the angle of arrival estimation points will not be limited to a single point, thus obtaining an reachability map, such as... Figure 2 As shown, the horizontal axis represents the clustering sequence, and the vertical axis represents the reachability distance. Each depression in the reachability graph represents a cluster.

[0085] Furthermore, for each cluster, the point with the highest density is selected as the estimated angle of arrival for that cluster. Represented as:

[0086]

[0087]

[0088]

[0089] Where X represents the set of data points in the cluster, X = {p1, ..., p...} i ,…,p j ,…,p n}, where subscripts i and j represent distinct data points, n represents the total number of data points, and ρ i This indicates the midpoint p of the cluster. i Density score, D ij p i and p j The distance between them.

[0090] According to the present invention, the entire process processes the single-captured incident signal in the digital domain, ensuring that the algorithm outperforms the traditional MUSIC algorithm in low signal-to-noise ratio environments, whether in additive white Gaussian noise (AWGN) channels or in actual multipath channels with Doppler effects.

[0091] Example

[0092] Example 1

[0093] A simulation experiment was conducted using a 4×4 uniform rectangular antenna array. The distance between any two adjacent antennas was d = λ / 2. Four signals were incident from the directions (φ1, θ1) = (-111.44, 30.45), (φ2, θ2) = (-135.89, 48.46), (φ3, θ3) = (-36.81, 56.69), and (φ4, θ4) = (-45.29, 52.34). The signal-to-noise ratio (SNR) was set to -10dB, and additive white Gaussian noise (AWGN) was added.

[0094] The angle of arrival can be estimated using the following method:

[0095] S1. Perform a two-dimensional beam scan in the sky to explore multiple potential angles of arrival;

[0096] S2. Create multiple overlapping virtual subarrays and perform digital beamforming in these directions to obtain multiple virtual subarray signals;

[0097] S3. The received signal is divided into multiple overlapping snapshots to obtain multiple possible angle of arrival estimates. The point sorting identification clustering structure method is used to perform cluster analysis on the estimates to obtain the final angle of arrival estimate.

[0098] In S1, such as Figure 3As shown, a uniform rectangular antenna array is used to scan the beam in all directions to obtain the power received in each direction, and the direction with power greater than a threshold is taken as the potential angle of arrival.

[0099] Power P received by the beam in a certain direction BeamScan Represented as:

[0100] P BeamScan =ω H *R xx *ω

[0101] R xx =AR S A H +σ 2 I

[0102] A=A(φ,θ)=[a(φ1,θ1),…,a(φ i ,θ i )…,a(φ P ,θ P )]

[0103] The threshold is set to 4.

[0104] In S2, such as Figure 4 As shown, a virtual subarray is created by virtually dividing a uniform rectangular antenna array into multiple subarrays, with each virtual subarray being 3×3 in size.

[0105] For each beamforming signal, the MUSIC method is used to estimate the angle of arrival, including the following sub-steps:

[0106] Obtain the autocorrelation matrix of the beamforming signal, denoted as R. yy =E[y j (t)y j H (t)];

[0107] For the autocorrelation matrix R yy Perform eigenvalue decomposition, denoted as R yy =UΛU H , thus obtaining the eigenvector matrix;

[0108] The signal is divided into two orthogonal subspaces by using eigenvectors: a signal subspace and a noise subspace. The signal of the virtual subarray is obtained based on the noise subspace using the MUSIC method.

[0109] The signal from this virtual subarray is transmitted through the MUSIC spatial spectrum function P. MUSIC express:

[0110]

[0111] μ p =sinφ p sinθ p v p =cosφ p sinθ p .

[0112] In S3, such as Figure 5 As shown, each signal is divided into 100 snapshots, and all potential angles of arrival are clustered. When the signal-to-noise ratio is low, the angle of arrival estimation points form an reachability map. The horizontal axis of the reachability map is the clustering sequence, and the vertical axis is the reachability distance. Each depression in the reachability map represents a cluster. For each cluster, the point with the highest density is selected as the angle of arrival estimate for that cluster.

[0113] Comparative Example 1

[0114] The same experiment as in Example 1 was conducted, except that the conventional MUSIC method was used.

[0115] RMSE was used as the evaluation index, and the results of Example 1 and Comparative Example 1 were compared, as shown in Table 1.

[0116] Table 1

[0117] RMSE Example 1 Comparative Example 1 <![CDATA[(φ1,θ1)]]> (0.0518,0.0492) (0.2704,0.2308) <![CDATA[(φ2,θ2)]]> (0.1090,0.0396) (12.3547,11.1985) <![CDATA[(φ3,θ3)]]> (0.1897,0.1920) (0.6801,0.5903) <![CDATA[(φ4,θ4)]]> (0.2975,0.1566) (15.7823,14.6329)

[0118] As can be seen from Table 1, the estimation accuracy of the method in Example 1 is significantly better than that of Comparative Example 1.

[0119] In the description of this invention, it should be noted that the terms "upper," "lower," "inner," "outer," "front," and "rear," etc., indicate the orientation or positional relationship based on the orientation or positional relationship in the working state of this invention, and are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention. Furthermore, the terms "first," "second," "third," and "fourth" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0120] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0121] The present invention has been described above with reference to preferred embodiments; however, these embodiments are merely exemplary and illustrative. Various substitutions and modifications can be made to the present invention based on these embodiments, all of which fall within the scope of protection of the present invention.

Claims

1. A method for accurate angle of arrival estimation based on OPTICS under low signal-to-noise ratio, characterized in that, The method includes the following steps: S1. Perform a two-dimensional beam scan in the sky to explore multiple potential angles of arrival; S2. Create multiple overlapping virtual subarrays and perform digital beamforming in these directions to obtain multiple virtual subarray signals; S3. The received signal is divided into multiple overlapping snapshots to obtain multiple possible angle of arrival estimates. The point sorting identification clustering structure method is used to perform cluster analysis on the estimates to obtain the final angle of arrival estimate.

2. The method for accurate angle of arrival estimation based on OPTICS under low signal-to-noise ratio conditions according to claim 1, characterized in that, In S1, a uniform rectangular antenna array is used to scan the beam in all directions to obtain the power received in each direction, and the direction with power greater than a threshold is taken as the potential angle of arrival.

3. The method for accurate angle of arrival estimation based on OPTICS under low signal-to-noise ratio conditions according to claim 1, characterized in that, Power P received by the beam in a certain direction BeamScan Represented as: P BeamScan =ω H *R xx *oh R xx =AR S A H +σ 2 I Where ω represents the weight vector of direction, H represents the conjugate transpose, and R... xx It is the array correlation matrix, A represents the rotation matrix, σ 2 R is the noise variance, I is the identity matrix, and R is the noise variance. s It is the autocorrelation matrix of the source signal S; The rotation matrix A is represented as: A=A(φ,θ)=[a(φ1,θ1),…,a(φ i ,i i )…,a(φ P ,i P )] Where φ represents the azimuth angle in that direction, θ represents the elevation angle in that direction, p represents the total number of directions of the incident signal, and a(φ) i ,θ i ) indicates direction (φ) i ,θ i The vector of the incident signal.

4. The method for accurate angle of arrival estimation based on OPTICS under low signal-to-noise ratio conditions according to claim 1, characterized in that, The threshold is set according to the following conditions: Condition 1: Peak value within a range of -3dB from the highest power peak; Condition 2: The maximum number of K highest peak values.

5. The method for accurate angle of arrival estimation based on OPTICS under low signal-to-noise ratio conditions according to claim 1, characterized in that, In S2, the virtual subarray is created by virtually dividing the uniform rectangular antenna array into multiple subarrays.

6. The method for accurate angle of arrival estimation based on OPTICS under low signal-to-noise ratio conditions according to claim 5, characterized in that, The size of the virtual subarray is (M-1)×(N-1).

7. The method for accurate angle of arrival estimation based on OPTICS under low signal-to-noise ratio conditions according to claim 1, characterized in that, In S2, for each beamforming signal, the MUSIC method is used to estimate the angle of arrival, including the following sub-steps: Obtain the autocorrelation matrix of the beamforming signal; For the autocorrelation matrix R yy Perform eigenvalue decomposition to obtain the eigenvector matrix; The signal is divided into two orthogonal subspaces by using eigenvectors: a signal subspace and a noise subspace. The signal of the virtual subarray is obtained based on the noise subspace using the MUSIC method.

8. The method for accurate angle of arrival estimation based on OPTICS under low signal-to-noise ratio conditions according to claim 1, characterized in that, In S3, each signal is divided into multiple signals to form a snapshot. The length of the segmented signal is shorter than that of the original signal, and there is overlap between the segmented signals.

9. The method for accurate angle of arrival estimation based on OPTICS under low signal-to-noise ratio conditions according to claim 1, characterized in that, Cluster all potential angles of arrival. When the signal-to-noise ratio is low, the angle of arrival estimation points form an reachability map. The horizontal axis of the reachability map is the cluster sequence, and the vertical axis is the reachability distance. Each depression in the reachability map represents a cluster. For each cluster, the point with the highest density is selected as the angle of arrival estimate for that cluster.

Images