Array signal direction estimation methods, electronic devices and positioning systems

Abstract

This application provides an array signal direction estimation method, an electronic device, and a positioning system; it relates to the field of communication technology. Given the problem of large errors or even failure in traditional signal direction estimation, the array signal direction estimation method provided in this application includes: obtaining the phase matrix of the target object at the current moment; obtaining the phase matrix based on the original positioning data collected by the positioning base station; calculating the second covariance matrix at the current moment based on the phase matrix at the current moment and the first covariance matrix at the previous moment; updating the second covariance matrix based on the gradient of the second covariance matrix to obtain a third covariance matrix; and calculating the angle information of the target object relative to the positioning base station using a direction estimation algorithm based on the third covariance matrix. The embodiments of this application can improve the accuracy of direction estimation for array signals.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to an array signal direction estimation method, electronic device and positioning system. Background Technology

[0002] In array signal processing, accurately estimating the Angle of Arrival (AOA) is crucial for systems such as wireless positioning, intelligent communication, and radar detection. Traditional direction estimation algorithms rely on constructing a covariance matrix from the received in-direction and orthogonal (IQ) component data, assuming the signal is stationary within the sampling period, and extracting the direction information of the array signal through the covariance matrix. However, traditional calculation methods are based on static, batch-acquired IQ data, estimating the covariance matrix in a single extraction. In complex scenarios such as rapid signal source movement, dynamic changes in the channel environment, low signal-to-noise ratio, and multipath interference, the distortion of the static covariance matrix leads to large direction estimation errors, or even failure. Summary of the Invention

[0003] According to various embodiments of this application, an array signal direction estimation method, an electronic device, and a positioning system are provided; these can improve the accuracy of direction estimation for array signals.

[0004] In a first aspect, this application provides an array signal direction estimation method, the method comprising:

[0005] Obtain the phase matrix of the target object at the current time; the phase matrix is ​​obtained based on the raw positioning data collected by the positioning base station; based on the phase matrix at the current time and the first covariance matrix at the previous time, calculate the second covariance matrix at the current time; based on the gradient of the second covariance matrix, update the second covariance matrix to obtain the third covariance matrix; based on the third covariance matrix, calculate the angle information of the target object relative to the positioning base station through a direction estimation algorithm.

[0006] The above method calculates the second covariance matrix at the current time based on the first covariance matrix at the previous time step, balancing the stability of historical data with the real-time performance of current data. Updates are performed based on the gradient of the second covariance matrix to suppress the impact of noise on direction estimation. Gradient-based updates quickly track changes in signal characteristics, avoiding errors caused by static assumptions and improving the accuracy and robustness of signal direction estimation in complex scenarios such as rapid movement and multipath interference. It also demonstrates strong ease of use and practicality.

[0007] In one possible implementation of the first aspect, obtaining the phase matrix of the target object at the current moment includes:

[0008] Obtain the original location data of the target object collected by the positioning base station at the current moment;

[0009] Based on the antenna array of the positioning base station, the phase matrix corresponding to the original positioning data is extracted.

[0010] In one possible implementation of the first aspect, the second covariance matrix at the current time is calculated based on the phase matrix at the current time and the first covariance matrix at the previous time, including:

[0011] Based on the phase matrix and the first covariance matrix, the second covariance matrix is ​​obtained through recursive calculation; the recursive calculation formula is expressed as follows:

[0012]

[0013] Among them, R t Let R be the second covariance matrix at the current time t. t-1 Let X be the first covariance matrix corresponding to the previous time step, α be the first weight coefficient, and X be the first covariance matrix. t For the phase matrix, It is the conjugate matrix of the phase matrix.

[0014] In one possible implementation of the first aspect, the second covariance matrix is ​​updated based on its gradient to obtain a third covariance matrix, including:

[0015] Based on the eigenvalues ​​of the second covariance matrix, a loss function for the second covariance matrix is ​​constructed;

[0016] Based on the eigenvalues ​​and eigenvectors of the second covariance, the gradient of the loss function with respect to the second covariance matrix is ​​calculated;

[0017] Based on the gradient, the second covariance matrix is ​​updated to obtain the third covariance matrix.

[0018] In one possible implementation of the first aspect, constructing the loss function for the second covariance matrix based on the eigenvalues ​​of the second covariance matrix includes:

[0019] Based on the maximum and minimum eigenvalues ​​of the second covariance matrix, the loss function is constructed; the expression of the constructed loss function is as follows:

[0020]

[0021] in, Let λ be the value of the loss function. max λ is the largest eigenvalue. minγ is the minimum eigenvalue, γ is the second weighting parameter, and R is the value of the second covariance matrix; the eigenvalues ​​include the maximum eigenvalue and the minimum eigenvalue.

[0022] In one possible implementation of the first aspect, calculating the gradient of the loss function with respect to the second covariance matrix based on the eigenvalues ​​and eigenvectors of the second covariance includes:

[0023] Based on the largest eigenvalue and its corresponding first eigenvector, and the smallest eigenvalue and its corresponding second eigenvector, the gradient of the loss function with respect to the second covariance matrix is ​​calculated; the expression for calculating the gradient is as follows:

[0024]

[0025] in, V is the gradient value. max The first eigenvector corresponding to the largest eigenvalue. V is the conjugate transpose of the first eigenvector. min The second eigenvector corresponding to the smallest eigenvalue. Let λ be the conjugate transpose of the second eigenvector. min λ is the smallest eigenvalue. max γ is the maximum eigenvalue, and γ is the second weighting parameter; the eigenvalues ​​include the minimum eigenvalue and the maximum eigenvalue.

[0026] In one possible implementation of the first aspect, updating the second covariance matrix based on the gradient to obtain the third covariance matrix includes:

[0027] Based on the gradient, determine the rate and direction of change of the second covariance matrix in space;

[0028] Based on the rate and direction of change, determine the minimum value of the loss function corresponding to the updated second covariance matrix;

[0029] The value of the second covariance matrix at the minimum value of the loss function is taken as the third covariance matrix; the expression for updating the second covariance matrix is ​​as follows:

[0030]

[0031] Among them, R t ' is the third covariance matrix; R t This is the second covariance matrix; η is the gradient corresponding to the second covariance matrix; η is the learning efficiency, which can take the value 0.01.

[0032] In one possible implementation of the first aspect, the method further includes:

[0033] If the norm difference between two consecutive update iterations of the covariance matrix is ​​within a preset convergence threshold, the updated second covariance matrix is ​​determined to be converged, and the third covariance matrix is ​​obtained.

[0034] In one possible implementation of the first aspect, the method further includes:

[0035] Based on the angle information of the target object relative to one or more of the positioning base stations, the positioning information of the target object is determined.

[0036] Secondly, this application provides an array signal direction estimation device, which includes:

[0037] The acquisition unit is used to acquire the phase matrix of the target object at the current time; the phase matrix is ​​obtained based on the raw positioning data collected by the positioning base station;

[0038] The first calculation unit is used to calculate the second covariance matrix at the current time based on the phase matrix at the current time and the first covariance matrix at the previous time.

[0039] An update unit is used to update the second covariance matrix based on the gradient of the second covariance matrix to obtain a third covariance matrix;

[0040] The second calculation unit is used to calculate the angle information of the target object relative to the positioning base station based on the third covariance matrix and through a direction estimation algorithm.

[0041] Thirdly, this application provides an electronic device including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method described in any one of the first aspects.

[0042] Fourthly, this application provides a positioning system, including a positioning base station and a server; the positioning base station is used to collect raw positioning data of a target object and send the raw positioning data to the server; the server is used to receive the raw positioning data sent by the positioning base station and implement the method described in any one of the first aspects based on the raw positioning data.

[0043] Fifthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described in any one of the first aspects.

[0044] Sixthly, this application provides a computer program product that, when run on a device, causes the device to perform the method described in any one of the first aspects above.

[0045] It is understood that the beneficial effects of the second to sixth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a schematic diagram of the overall architecture of the orientation estimation system provided in the embodiments of this application;

[0048] Figure 2 A schematic diagram illustrating the implementation process of the array signal direction estimation method provided in this application embodiment;

[0049] Figure 3 This is a schematic diagram of the overall framework provided for an embodiment of this application;

[0050] Figure 4 This is a schematic diagram illustrating the principle of collecting raw positioning data provided in an embodiment of this application;

[0051] Figure 5 This is a schematic diagram of the array signal direction estimation device provided in an embodiment of this application;

[0052] Figure 6 The present application provides a schematic diagram of the structure of an electronic device. Detailed Implementation

[0053] The embodiments of the technical solution of this application will now be described in detail with reference to the accompanying drawings. These embodiments are only used to more clearly illustrate the technical solution of this application and are therefore merely examples, and should not be used to limit the scope of protection of this application.

[0054] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0055] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0056] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0057] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0058] More and more factories are demanding high-precision, low-latency real-time positioning for their warehousing and logistics processes. In complex industrial environments (such as high racks, stacked pallets, and dense metal structures), traditional radio frequency or signal strength-based positioning solutions are easily affected by obstructions, multipath propagation, or signal attenuation, and their positioning accuracy cannot meet the requirements of scenarios such as automated guided forklifts, intelligent picking robots, and precise material tracking.

[0059] In related technologies, factories deploy Angle of Arrival (AOA) positioning systems. Traditional signal estimation methods typically rely on static, batch-collected data, extracting it all at once to estimate the covariance matrix and thus the signal direction. However, in complex scenarios such as rapidly moving signal sources, dynamically changing channel environments, and multipath propagation, the static covariance matrix obtained by averaging using a fixed window exhibits lag and distortion due to the continuous movement of handling robots, forklifts, and material pallets, severely impacting the accuracy of signal angle estimation. Furthermore, in complex scenarios, multipath effects lead to feature overlap, causing data ambiguity or misalignment in the input direction estimation algorithm, affecting the accuracy of signal direction determination. Additionally, for non-stationary targets or abrupt changes (such as personnel crossing or forklifts entering or leaving an area), angle jumps and misjudgments of target objects are prone to occur.

[0060] To address the above technical issues, this application provides an array signal direction estimation method. When a new frame of raw positioning data of a target object is received, the corresponding covariance matrix is ​​updated using a weighted recursive method. Based on the gradient descent method, the main direction features of the signal are continuously strengthened and noise interference is suppressed, so that the covariance matrix is ​​optimized to a structure that is more conducive to direction estimation. In complex environments, this improves the accuracy of target object direction estimation.

[0061] The following example illustrates the specific implementation process of this array signal direction estimation method.

[0062] Please see Figure 1 , Figure 1 This is a schematic diagram of the overall architecture of the direction estimation system provided in the embodiments of this application; as shown below. Figure 1 As shown, this orientation estimation system may include positioning base stations and electronic devices (such as servers); there may be multiple positioning base stations acting as signal receivers (RX), and the target object (such as an electronic tag) acting as a signal transmitter (TX). Specifically, the positioning base stations receive signals emitted by the target object in real time based on an antenna array. The positioning base stations transmit the received signals to the electronic devices, which calculate the orientation information of the target object relative to the base station, such as angle 1 and angle 2, based on the signals received by each antenna; and then determine the positioning information of the target object based on angle 1 and angle 2.

[0063] Based on the above system architecture, the implementation process of the array signal direction estimation method will be further described below.

[0064] Please see Figure 2 , Figure 2 This is a schematic diagram illustrating the implementation flow of the array signal direction estimation method provided in this application embodiment. The execution entity of this method can be... Figure 1 The electronic device in the middle, such as a server, etc. Figure 2 As shown, the method may include the following steps:

[0065] S201, Obtain the phase matrix of the target object at the current moment; the phase matrix is ​​obtained based on the raw positioning data collected by the positioning base station.

[0066] In this embodiment of the application, the positioning base station collects the original positioning data of the target object through an array antenna, such as in-phase and quadrature (IQ) data; the electronic device acquires the original positioning data sent by the positioning base station.

[0067] For example, the raw positioning data can be a radio frequency signal or an ultrasonic signal; the phase matrix is ​​a multi-dimensional data structure used to describe the signal phase relationship between the target object and the positioning base station, and is calculated based on the raw positioning data. Figure 3As shown, the positioning base station can be an AOA positioning base station. The electronic device can simultaneously receive IQ data collected by multiple AOA positioning base stations, and calculate the corresponding phase matrix for the IQ data of each AOA positioning base station.

[0068] By acquiring the phase matrix of the target object, the phase information can record the direct path and the reflected path in complex environments (such as indoor multipath propagation), thereby improving the robustness of direction estimation. At the same time, due to the sensitivity of phase to the signal propagation path, it is suitable for high-precision direction estimation and supports the tracking of dynamic target objects.

[0069] In some embodiments, obtaining the phase matrix of the target object at the current moment includes:

[0070] Obtain the raw positioning data of the target object collected by the positioning base station at the current moment; extract the phase matrix corresponding to the raw positioning data based on the antenna array of the positioning base station.

[0071] For example, each time frame, the positioning base station collects a new frame of raw positioning data, such as IQ data. The IQ data can be sampled data of the raw voltage or current waveforms received by the positioning base station. When the electronic device receives a new frame of raw positioning data, it extracts the phase matrix corresponding to the raw positioning data based on the number of antenna arrays of the positioning base station. For example, if the antenna array of the positioning base station is an 8*8 matrix, the 8*8 phase matrix corresponding to each antenna is calculated based on the collected IQ data.

[0072] like Figure 4 The antenna plate shown in (a) has 8 antenna positions, forming an antenna array symmetrically distributed around the center. It receives signals through the antennas and generates an 8*8 phase matrix based on the received signals. Each element in the phase matrix represents the channel response of the signal to a receiving antenna. The corresponding phase matrix is ​​generated by measuring the signal transmission characteristics (such as amplitude and phase) of the antenna.

[0073] Specifically, based on the in-phase and quadrature components in the IQ data, the corresponding phase is calculated, such as the phase... I represents the sampled in-phase component, Q represents the sampled quadrature component, and i is a complex number. For example, based on the raw positioning data collected by the m×n antenna array of each positioning base station, the following phase matrix X is calculated:

[0074]

[0075] Let m be the phase corresponding to the raw positioning data collected by one of the antennas in the antenna array, where m can be equal to n.

[0076] Among them, such as Figure 4 The principle of time-division sampling of IQ data, as shown in (b), involves the positioning base station collecting IQ data and transmitting it to electronic devices (such as the positioning engine module of a server). Figure 4 As shown in (b), the positioning base station can use two different durations (such as 1 microsecond and 2 microseconds) of switching and sampling time slots to adapt to different application scenarios or signal characteristics; for example, shorter time slots (such as 1 microsecond) can be used in scenarios with high time requirements and rapid signal changes, while longer time slots (such as 2 microseconds) can be used in scenarios with weak signals or high noise to obtain more accurate sampling data.

[0077] like Figure 4 As shown in (b), the transmission process of the positioning base station to the electronic device, sending the angle of arrival, is continuous, providing a constant signal input to the electronic device (receiver). The positioning base station also sets a guard interval, such as 4 microseconds, during signal reception to avoid interference between different signals or operations. For example, during switching receiving channels or other conversion operations, there may be transient responses or interference signals. Setting a guard interval provides a buffer time for these unstable states, ensuring the accuracy of subsequent sampling data.

[0078] For example, the positioning base station also sets a reference period, such as 8 microseconds, to provide a reference signal or synchronization information, enabling the positioning base station to better understand and process subsequent sampled data. The reference period can serve as a reference standard for time or signal strength to calibrate and adjust the parameters of the positioning base station.

[0079] Switching slots (i.e., switching time slots) are used to switch receiving channels or perform other related operations. In angle-of-arrival (AOA) measurements, multiple receiving channels may be needed to receive signals from different directions. Switching time slots enable the switching between these channels so that the positioning base station can receive signals from each channel sequentially.

[0080] The sampling slot (i.e., the sampling time slot) is the actual time period during which the signal is sampled; within the sampling time slot, the positioning base station samples the received signal to obtain IQ data; Figure 4 (b) shows multiple sampling time slots (such as sample slot 1, sample slot 2, etc.), each corresponding to sequential signal sampling. The positioning base station acquires multiple sets of IQ data samples through multiple sampling time slots for subsequent signal processing and analysis. Different sampling time slots can be selected based on the accuracy requirements of the actual application scenario. More sampling time slots help improve the accuracy and reliability of the measurement.

[0081] Accordingly, when acquiring raw positioning data from multiple positioning base stations, the time of the multiple positioning base stations is synchronized. In addition to acquiring raw positioning data from multiple positioning base stations, the raw positioning data can also be filtered to improve the signal-to-noise ratio of the raw positioning signal.

[0082] S202, based on the phase matrix at the current time and the first covariance matrix at the previous time, calculate the second covariance matrix at the current time.

[0083] In some embodiments, the first covariance matrix of the previous time step is a signal statistical feature matrix estimated based on historical data; the second covariance matrix is ​​the covariance matrix of the current time step updated by combining the phase matrix of the previous time step, used for dynamically tracking the target object. By using a weighted recursive method to update the covariance matrix, historical information and the state of the target object at the current time step are balanced, improving the adaptability of the positioning system to dynamic target objects.

[0084] For example, the covariance matrix is ​​the mathematical expectation of the phase matrix and its own conjugate transpose, i.e., the covariance matrix R = E[XX]. H R is the covariance matrix, X is the phase matrix, and Xp is the phase matrix. H This is the conjugate transpose of the phase matrix. For example, based on the above phase matrix, the covariance matrix is ​​calculated as follows:

[0085]

[0086] Where, r m,n The elements in the covariance matrix represent the correlation, or covariance, between the original positioning signals received by different antennas.

[0087] Accordingly, the first covariance matrix of the previous time step is calculated based on the phase matrix corresponding to the original positioning data collected at the previous time step and the covariance matrix of the time step before that. Based on the covariance matrix calculated from the phase matrix at the current time step, and combined with the first covariance matrix of the previous time step, the second covariance matrix of the current time step is obtained.

[0088] Specifically, the covariance matrix calculated from the phase matrix at the current moment is combined with the first covariance matrix at the previous moment to obtain the second covariance matrix at the current moment through a weighted recursive calculation. Upon initial system startup, an initial covariance matrix can be preset based on the characteristics of the operating environment and noise levels, which will be used for subsequent calculations of the weighted recursive covariance matrix.

[0089] In some embodiments, the second covariance matrix at the current time is calculated based on the phase matrix at the current time and the first covariance matrix at the previous time, including:

[0090] Based on the phase matrix and the first covariance matrix, the second covariance matrix is ​​obtained through recursive calculation; the recursive calculation formula is expressed as follows:

[0091]

[0092] Among them, Rt Let R be the second covariance matrix at the current time t. t-1 Let X be the first covariance matrix corresponding to the previous time step, α be the first weight coefficient, and X be the first covariance matrix. t For the phase matrix, It is the conjugate matrix of the phase matrix.

[0093] For example, the first weighting coefficient is used to control the depth of memory, i.e. the importance of historical data, and is used to improve the tracking capability and stability of the positioning system. It can be adjusted according to the actual application scenario, and the value range is 0 to 1. For example, for application scenarios with stable or slowly changing signal environments, the value of the first weighting coefficient can be larger (e.g., 0.95 to 0.99), making the subsequent direction estimation more robust to noise and transient interference. When the signal environment changes rapidly and is dynamic, the value of the first weighting coefficient can be smaller (e.g., 0.5 to 0.7), so that the rapid movement of the signal source and environmental changes can be tracked quickly.

[0094] For example, the value of the first weighting coefficient can be adaptively adjusted based on the actual application environment. When the positioning system detects a sudden change in the environment, it can automatically reduce the value of the first weighting coefficient to enhance the tracking capability; when the system is in a stable state, it can automatically increase the value to improve the accuracy and stability of signal direction estimation.

[0095] S203, based on the gradient of the second covariance matrix, update the second covariance matrix to obtain the third covariance matrix.

[0096] In some embodiments, the second covariance matrix is ​​the matrix to be optimized or the matrix in the iterative update process, which contains all the signal information at the current time, such as the main peak features and noise features of the signal; the gradient is calculated based on the eigenvalues ​​and eigenvectors of the second covariance matrix, and is used to indicate the iterative update direction of each element in the second covariance matrix, so that the signal information represented by the second covariance matrix is ​​updated towards the structural update of the signal features with prominent main peaks and flat noise.

[0097] For example, the second covariance matrix is ​​decomposed into eigenvalues ​​and eigenvectors. The gradient of the second covariance matrix is ​​calculated based on the eigenvalues ​​and eigenvectors. The elements of the second covariance matrix are iteratively updated based on the gradient values. With each updated second covariance matrix, eigenvalue decomposition is performed again to obtain eigenvalues ​​and eigenvectors. The corresponding gradient is calculated again based on the eigenvalues ​​and eigenvectors. This process continues until the second covariance matrices of two consecutive updates satisfy a preset convergence condition, such as the difference norm of the two matrices being less than a preset convergence threshold. The optimized third covariance matrix is ​​then obtained.

[0098] In some embodiments, the second covariance matrix is ​​updated based on its gradient to obtain the third covariance matrix, including:

[0099] Based on the eigenvalues ​​of the second covariance matrix, a loss function for the second covariance matrix is ​​constructed; based on the eigenvalues ​​and eigenvectors of the second covariance matrix, the gradient of the loss function relative to the second covariance matrix is ​​calculated; based on the gradient, the second covariance matrix is ​​updated to obtain the third covariance matrix.

[0100] For example, eigenvalues ​​and eigenvectors are obtained by performing eigendecomposition on the second covariance matrix. The eigendecomposition expression is R = UΛU H Where R is the second covariance matrix, U is the eigenvector matrix, and Λ is the eigenvalue diagonal matrix. H It is the conjugate transpose of the eigenvector matrix.

[0101] For example, since the acquired phase matrix is ​​affected by noise, a loss function for the second covariance matrix is ​​constructed to strengthen the principal eigenvalues ​​and impose conditional constraints, thereby adjusting the balance between principal energy and numerical stability. The loss function enhances the influence of the principal eigenvalues ​​by maximizing them and incorporating the ratio of the maximum to the minimum eigenvalues. For instance, the loss function for the second covariance matrix is ​​calculated based on the maximum eigenvalue (i.e., the principal eigenvalue) and the ratio of the maximum to the minimum eigenvalue.

[0102] For example, eigenvalues ​​and eigenvectors are obtained by eigenvalue decomposition of the second covariance matrix; the gradient of the second covariance is calculated based on the eigenvalues ​​and eigenvectors, which represents the rate of change of the loss function with respect to the matrix elements at the current second covariance matrix and points in the direction of the fastest growth of the loss function; during the optimization process, the second covariance matrix is ​​updated along the opposite direction of the gradient to reduce the value of the loss function.

[0103] For example, based on the gradient descent method, the gradient at the current second covariance matrix is ​​continuously calculated, and the second covariance matrix is ​​updated along the opposite direction of the gradient according to the learning rate to determine the step size, so that the value of the loss function gradually decreases, the second covariance matrix is ​​gradually optimized, and finally the third covariance matrix that satisfies the convergence condition is obtained.

[0104] In some embodiments, a loss function for the second covariance matrix is ​​constructed based on the eigenvalues ​​of the second covariance matrix, including:

[0105] Based on the maximum and minimum eigenvalues ​​of the second covariance matrix, a loss function is constructed; the expression of the constructed loss function is as follows:

[0106]

[0107] in, Let λ be the value of the loss function. max λ is the largest eigenvalue. min γ is the minimum eigenvalue, γ is the second weighting parameter, and R is the value of the second covariance matrix; the eigenvalues ​​include the maximum eigenvalue and the minimum eigenvalue.

[0108] For example, the largest eigenvalue is used to represent the variance of the data in the second covariance matrix along the corresponding eigenvector direction. The direction corresponding to the largest eigenvalue is usually the direction of the greatest data variation, i.e., the principal direction; the direction corresponding to the smallest eigenvalue has relatively smaller data variation. The second weight parameter is used to adjust the balance between principal energy and numerical stability, thereby ensuring the numerical stability of the calculation process while ensuring the enhancement of principal eigenvalues ​​(principal energy directions).

[0109] Correspondingly, the first term in the formula, due to the preceding negative sign, maximizes the principal eigenvalue during the optimization process, highlighting the main directional information of the data; the second term in the formula, through the ratio of the maximum eigenvalue to the minimum eigenvalue, considers the overall distribution of the data, strengthens the principal eigenvalue while maintaining the stability of the value, and avoids extreme cases affecting the calculation results.

[0110] In some embodiments, the gradient of the loss function with respect to the second covariance matrix is ​​calculated based on the eigenvalues ​​and eigenvectors of the second covariance, including:

[0111] Based on the largest eigenvalue and its corresponding first eigenvector, and the smallest eigenvalue and its corresponding second eigenvector, the gradient of the loss function with respect to the second covariance matrix is ​​calculated; the expression for calculating the gradient is as follows:

[0112]

[0113] in, V is the gradient value. max The first eigenvector corresponding to the largest eigenvalue. V is the conjugate transpose of the first eigenvector. min The second eigenvector corresponding to the smallest eigenvalue. Let λ be the conjugate transpose of the second eigenvector. min λ is the smallest eigenvalue. max γ is the largest eigenvalue, and γ is the second weighting parameter; the eigenvalues ​​include the smallest eigenvalue and the largest eigenvalue.

[0114] For example, the first eigenvector corresponding to the largest eigenvalue defines the direction of data change, pointing to the direction of the largest data change; the second eigenvector corresponding to the smallest eigenvalue represents the direction of smaller data change, and the second weight parameter is used to adjust the balance between principal energy and numerical stability.

[0115] For example, the above gradient calculation formula can be obtained by taking the derivative of the loss function with respect to the second covariance matrix; or by using the chain rule of matrix differentiation based on the intrinsic relationship between eigenvalues ​​and eigenvectors and the second covariance matrix.

[0116] In some embodiments, the second covariance matrix is ​​updated based on the gradient to obtain the third covariance matrix, including:

[0117] Based on the gradient, the rate and direction of change of the second covariance matrix in space are determined; based on the rate and direction of change, the minimum value of the loss function corresponding to the updated second covariance matrix is ​​determined; the value of the second covariance matrix corresponding to the minimum value of the loss function is taken as the third covariance matrix; the expression for updating the second covariance matrix is ​​as follows:

[0118]

[0119] Among them, R t ' is the third covariance matrix; R t This is the second covariance matrix; η is the gradient corresponding to the second covariance matrix; η is the learning efficiency, which can take the value 0.01.

[0120] For example, the gradient indicates the direction in which the loss function grows fastest, and thus the opposite direction of the gradient points to the direction in which the loss function decreases fastest. The magnitude (norm) of the gradient determines the rate of change of the loss function; the larger the gradient, the faster the loss function decreases. The direction of the gradient is the direction in which the loss function changes. Based on the direction and rate of change, the second covariance is iteratively updated according to the learning rate in the direction of the expected minimum value of the loss function, resulting in a smaller loss function value, such as the minimum value of the loss function; or the gradient is close to zero, or the second covariance matrix hardly changes, or the change is very small. By repeating the above update process multiple times, the value of the second covariance matrix corresponding to when the loss function converges to the minimum value is used as the optimized third covariance matrix.

[0121] In some embodiments, the method further includes:

[0122] If the norm difference between two consecutive update iterations of the covariance matrix is ​​within a preset convergence threshold, the updated second covariance matrix is ​​determined to be converged, and the third covariance matrix is ​​obtained.

[0123] For example, a convergence threshold is used to determine whether the update process of the second covariance matrix can be stopped. This is achieved by comparing the degree of change in the covariance matrix obtained from two adjacent iterations. When the change is sufficiently small, it indicates that the covariance matrix has stabilized, reaching a relatively optimized state, and the iteration process ends. The convergence judgment can be expressed as follows:

[0124] ||R t(i+1)-R t (i)||2<ε;

[0125] Among them, R t (i+1) and R t (i) are the second covariance matrices calculated in the (i+1)th and i-th iterations during the update process at the current time t, respectively; ||·||2 represents the 2-norm, which is used to measure the degree of change of the covariance matrix between two iterations; ε is the convergence threshold, which can be set to 0.01. When the 2-norm between the covariance matrices of two iterations is less than the convergence threshold, it is determined that the updated second covariance matrix has converged, that is, the third covariance matrix is ​​obtained.

[0126] S204, based on the third covariance matrix, calculates the angle information of the target object relative to the positioning base station through a direction estimation algorithm.

[0127] In some embodiments, eigenvalue decomposition is performed on the third covariance to obtain an eigenvector matrix and an eigenvalue diagonal matrix, thereby obtaining eigenvalues ​​and their corresponding eigenvectors. Based on the magnitude of the eigenvalues, a signal subspace and a noise subspace are determined; for example, the eigenvectors corresponding to larger eigenvalues ​​constitute the signal subspace, and the eigenvectors corresponding to smaller eigenvalues ​​constitute the noise subspace. The eigenvectors corresponding to the noise subspace are extracted to obtain the noise subspace matrix V. Based on the eigenvectors and steering vectors of the noise subspace in the third covariance matrix, the angle information of the target object relative to the positioning base station is calculated.

[0128] For example, the direction estimation algorithm for calculating this angle information can be a Multiple Signal Classification (MUSIC) algorithm, and the expression for calculating the angle information is as follows:

[0129]

[0130] Among them, P music (θ) represents the spectral function value of the angle of the target object relative to the positioning base station. By calculating this function value at different angles θ, the angle that results in a larger function value is found; this angle represents the angle information of the target object relative to the positioning base station, such as the elevation and azimuth angles. a(θ) is the steering vector, related to the angle θ; different angles correspond to different steering vectors, used to indicate the signal propagation characteristics from the positioning base station to the target object in different directions. V is the eigenvector matrix of the noise subspace, obtained by the eigenvalue decomposition of the third covariance matrix, i.e., the eigenvectors corresponding to the smallest eigenvalues ​​form the noise subspace matrix V; V H Let a(θ) be the conjugate transpose of the noise subspace matrix. H It is the conjugate transpose of the guide vector.

[0131] The steering vector is determined based on the structure of the antenna array of the positioning system and the signal propagation model. For example, for a uniform linear array, the steering vector can be determined based on the spacing between antenna elements, the wavelength of the signal, the number of antenna elements, and the angle to be estimated (such as the elevation angle or azimuth angle). Different angles correspond to different steering vectors.

[0132] For example, during the calculation process, within a certain angular range (e.g., pitch angle range of -90° to 90°, azimuth angle range of 0° to 360°), P is calculated for each angle θ at certain angular intervals (e.g., 1°). music The value of (θ). Calculate P for all angles. music After determining the (θ) value, the angle information of the target object relative to the positioning base station is determined by finding the peak value in the angle spectrum. The angle θ corresponding to the peak value is the estimated target direction of the signal. Generally, the higher the peak value, the greater the probability that the angle is the target direction.

[0133] In some embodiments, the method further includes: determining the positioning information of the target object based on the angle information of the target object relative to multiple positioning base stations.

[0134] For example, when the coordinates of the positioning base stations are known, based on the angle information of the target object relative to each positioning base station and the position information of the positioning base stations, a system of equations is listed through geometric relationships. By solving the system of equations, the positioning information of the target object can be obtained.

[0135] Through the embodiments of this application, the covariance matrix is ​​updated based on recursive updates of the current covariance matrix and gradient optimization. The effectiveness of the covariance matrix is ​​improved through structural adjustment, preserving historical features while quickly adapting to new changes. This solves the problem of insufficient adaptation of static matrices to rapidly changing signals. Furthermore, through optimization, the signal direction features (such as the maximum eigenvalue) are actively enhanced, and the condition number is suppressed, improving the resolution of the direction estimation algorithm under complex channels. This solves the problems of lag, instability, and distortion in traditional methods, and improves the stability and accuracy of direction estimation in practical applications.

[0136] This application employs a combination of dynamic recursive updates and gradient optimization to iteratively and adaptively construct the covariance matrix. By using exponential weighting, the covariance matrix is ​​iteratively updated after each frame of original positioning data input, enhancing its responsiveness to changes in the main signal. By constructing an eigenvalue-based loss function and adjusting the covariance matrix structure through gradient descent, it becomes more conducive to subspace extraction and interference suppression. This application achieves fine-grained control and iterative optimization of the covariance structure of the original positioning data, thereby supporting the orientation estimation algorithm to maintain good direction estimation performance in dynamic environments, meeting the business requirements of high accuracy, strong robustness, and low latency. It maintains high angular resolution and strong anti-interference capabilities even in complex environments. By introducing conditional regularization constraints into the loss function, covariance matrix degradation is avoided, improving subspace recognition capabilities. It exhibits strong smoothing effects for dynamic trajectories, making it less prone to abrupt changes in trajectory estimation.

[0137] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0138] Corresponding to the array signal direction estimation method described in the above embodiments, Figure 5 A structural block diagram of an array signal direction estimation device provided in an embodiment of this application is shown. For ease of explanation, only the parts related to the embodiments of this application are shown.

[0139] Reference Figure 5 The device includes:

[0140] The acquisition unit 51 is used to acquire the phase matrix of the target object at the current time; the phase matrix is ​​obtained based on the raw positioning data collected by the positioning base station;

[0141] The first calculation unit 52 is used to calculate the second covariance matrix at the current time based on the phase matrix at the current time and the first covariance matrix at the previous time.

[0142] Update unit 53 is used to update the second covariance matrix based on the gradient of the second covariance matrix to obtain the third covariance matrix;

[0143] The second calculation unit 54 is used to calculate the angle information of the target object relative to the positioning base station based on the third covariance matrix and through a direction estimation algorithm.

[0144] Each of the above units is used to implement the steps in the above method embodiments.

[0145] Figure 6 A schematic diagram of the hardware structure of electronic device 6 is shown.

[0146] like Figure 6 As shown, the electronic device 6 of this embodiment includes: at least one processor 61 ( Figure 6 (Only one is shown in the image), memory 62, which stores a computer program 63 that can run on the processor 61. When the processor 61 executes the computer program 63, it implements the steps in the above method embodiments, for example... Figure 2 S201 to S204 are shown. Alternatively, when the processor 61 executes the computer program 63, it implements the functions of each module / unit in the above-described device embodiments.

[0147] It is understood that the structures illustrated in the embodiments of this application do not constitute a specific limitation on the electronic device 6. In other embodiments of this application, the electronic device 6 may include more or fewer components than illustrated, or combine some components, or split some components, or have different component arrangements. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0148] The electronic device 6 may include, but is not limited to, a processor 61 and a memory 62. Those skilled in the art will understand that... Figure 6 This is merely an example of electronic device 6 and does not constitute a limitation on electronic device 6. It may include more or fewer components than shown, or combine certain components, or different components. For example, the server may also include input sending devices, network access devices, buses, etc.

[0149] The processor 61 mentioned above can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0150] The processor 61 may also include a memory for storing instructions and data. In some embodiments, the memory in the processor 61 is a cache memory. This memory can store instructions or data that the processor 61 has just used or that are used repeatedly. If the processor 61 needs to use the instruction or data again, it can directly retrieve it from the memory. This avoids repeated accesses, reduces the waiting time of the processor 61, and thus improves the efficiency of the system.

[0151] In some embodiments, the aforementioned memory 62 may be an internal storage unit of the electronic device 6, such as a hard disk or memory. The memory 62 may also be an external storage device of the electronic device 6, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the electronic device 6. Furthermore, the memory 62 may include both internal and external storage units of the electronic device 6. The memory 62 is used to store operating systems, applications, bootloaders, data, and other programs, such as program code for computer programs. The memory 62 can also be used to temporarily store data that has been sent or is about to be sent.

[0152] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0153] It should be noted that the structure of the above-mentioned electronic device is only illustrative and may include other physical structures depending on the application scenario. The physical structure of the electronic device is not limited here.

[0154] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0155] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps described in the various method embodiments above.

[0156] This application provides a computer program product that, when run on a server, enables the server to execute the steps described in the above-described method embodiments.

[0157] If the integrated modules / units are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0158] The electronic devices, computer storage media, and computer program products provided in the embodiments of this application are all used to execute the methods provided above. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects corresponding to the methods provided above, and will not be repeated here.

[0159] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0160] It should be understood that the above description is merely to help those skilled in the art better understand the embodiments of this application, and is not intended to limit the scope of the embodiments of this application. Based on the examples given above, those skilled in the art can obviously make various equivalent modifications or changes. For example, some steps in the various embodiments of the above detection method may be unnecessary, or new steps may be added. Alternatively, any combination of two or more of the above embodiments may be used. Such modifications, changes, or combinations also fall within the scope of the embodiments of this application.

[0161] It should also be understood that the methods, situations, categories, and classifications of embodiments in this application are for the convenience of description only and should not constitute a special limitation. Various methods, categories, situations, and features in embodiments can be combined without contradiction.

[0162] It should also be understood that, in the various embodiments of this application, unless otherwise specified or in case of logical conflict, the terms and / or descriptions between different embodiments are consistent and can be referenced by each other, and the technical features in different embodiments can be combined to form new embodiments according to their inherent logical relationships.

[0163] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0164] In the embodiments provided in this application, it should be understood that the disclosed apparatus / network devices and methods can be implemented in other ways. For example, the apparatus / network device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0165] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0166] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

[0167] Finally, it should be noted that the above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for estimating the direction of an array signal, characterized in that, The method includes: Obtain the phase matrix of the target object at the current moment; the phase matrix is ​​obtained based on the raw positioning data collected by the positioning base station; Based on the phase matrix at the current time and the first covariance matrix at the previous time, the second covariance matrix at the current time is calculated. Based on the gradient of the second covariance matrix, the second covariance matrix is ​​updated to obtain the third covariance matrix; Based on the third covariance matrix, the angle information of the target object relative to the positioning base station is calculated using a direction estimation algorithm; Based on the gradient of the second covariance matrix, the second covariance matrix is ​​updated to obtain the third covariance matrix, which includes: Based on the eigenvalues ​​of the second covariance matrix, a loss function for the second covariance matrix is ​​constructed; Based on the eigenvalues ​​and eigenvectors of the second covariance, the gradient of the loss function with respect to the second covariance matrix is ​​calculated; Based on the gradient, the second covariance matrix is ​​updated to obtain the third covariance matrix; The step of constructing a loss function for the second covariance matrix based on its eigenvalues ​​includes: The loss function is constructed based on the maximum and minimum eigenvalues ​​of the second covariance matrix.

2. The method according to claim 1, characterized in that, Obtain the phase matrix of the target object at the current moment, including: Obtain the original location data of the target object collected by the positioning base station at the current moment; Based on the antenna array of the positioning base station, the phase matrix corresponding to the original positioning data is extracted.

3. The method according to claim 1, characterized in that, Based on the phase matrix at the current time and the first covariance matrix at the previous time, the second covariance matrix at the current time is calculated, including: Based on the phase matrix and the first covariance matrix, the second covariance matrix is ​​obtained through recursive calculation; the recursive calculation formula is expressed as follows: ； in, Let be the second covariance matrix corresponding to the current time t. Let be the first covariance matrix corresponding to the previous time step. As the first weighting coefficient, For the phase matrix, It is the conjugate matrix of the phase matrix.

4. The method according to claim 1, characterized in that, The constructed loss function expression is as follows: ； in, The value of the loss function. The largest eigenvalue, It is the smallest eigenvalue. R is the second weighting parameter, and R is the value of the second covariance matrix; the eigenvalues ​​include the maximum eigenvalue and the minimum eigenvalue.

5. The method according to claim 1, characterized in that, Based on the eigenvalues ​​and eigenvectors of the second covariance, the gradient of the loss function with respect to the second covariance matrix is ​​calculated, including: Based on the largest eigenvalue and its corresponding first eigenvector, and the smallest eigenvalue and its corresponding second eigenvector, the gradient of the loss function with respect to the second covariance matrix is ​​calculated; the expression for calculating the gradient is as follows: ； in, The gradient value, The first eigenvector corresponding to the largest eigenvalue. Let be the conjugate transpose of the first eigenvector. The second eigenvector corresponding to the smallest eigenvalue. Let be the conjugate transpose of the second eigenvector. It is the smallest eigenvalue. The largest eigenvalue, The second weighting parameter is used; the eigenvalues ​​include the minimum eigenvalue and the maximum eigenvalue.

6. The method according to claim 1, characterized in that, Based on the gradient, the second covariance matrix is ​​updated to obtain the third covariance matrix, including: Based on the gradient, determine the rate and direction of change of the second covariance matrix in space; Based on the rate and direction of change, determine the minimum value of the loss function corresponding to the updated second covariance matrix; The value of the second covariance matrix at the minimum value of the loss function is taken as the third covariance matrix; the expression for updating the second covariance matrix is ​​as follows: ； in, This is the third covariance matrix; This is the second covariance matrix; The gradient corresponding to the second covariance matrix; For learning efficiency.

7. The method according to claim 6, characterized in that, The method further includes: If the norm difference between two consecutive update iterations of the covariance matrix is ​​within a preset convergence threshold, the updated second covariance matrix is ​​determined to be converged, and the third covariance matrix is ​​obtained.

8. An electronic device, characterized in that, It includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method of any one of claims 1 to 7.

9. A positioning system, characterized in that, The positioning system includes a positioning base station and a server; The positioning base station is used to collect the original positioning data of the target object and send the original positioning data to the server. The server is configured to receive raw positioning data sent by the positioning base station and execute the method of any one of claims 1 to 7 based on the raw positioning data.

Images