A method and system for measuring the angle of arrival of a Bluetooth signal
By solving the IQ data and correcting the covariance matrix of Bluetooth signals, the problem of low signal strength positioning accuracy and insufficient angle estimation of AoA positioning under the influence of multipath coherence in traditional Bluetooth positioning technology is solved, and a higher signal angle estimation is achieved Accuracy and lower array design complexity.
Patent Information
- Application Number
- CN202210998720.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-08-19
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2042-08-19
AI Technical Summary
Traditional Bluetooth positioning technology relies on signal strength, has low accuracy and complex base station deployment. AoA positioning is insufficient angle estimation accuracy under the influence of indoor multipath coherence, and traditional solution coherence algorithms have limited performance, which increases array cost and difficulty.
By solving the IQ data of the array antenna, dewinding, phase compensation, covariance matrix correction and eigenvalue decomposition, the rank loss of the covariance matrix is reduced, the signal angle estimation accuracy is improved, and the array arrangement requirements are reduced.
Effectively suppress multipath coherence influence in indoor environments, improve the accuracy of wave angle estimation of Bluetooth signals, reduce positioning errors, simplify array design, and reduce costs.
Smart Images

Figure CN115267656B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of communication technology, and in particular to a method and system for measuring the angle of arrival of a Bluetooth signal. Background Art
[0002] Among many positioning technologies, Bluetooth positioning technology has unique advantages in terms of cost, low power consumption and promotion efficiency. Traditional Bluetooth positioning technology mainly relies on received signal strength (RSS). One type uses related algorithms to convert RSS into distance data between beacons and base stations, and the other is to establish a signal strength fingerprint library for indoor environments and perform real-time matching during positioning. The principle of signal strength positioning is as follows: Figure 1 shown.
[0003] The technology that achieves positioning based on signal strength has complex base station deployment and requires more than three base stations to achieve positioning. Moreover, the circles with the three base stations as the center cannot intersect at one point, so this positioning method has low accuracy, usually 2 to 4 meters.
[0004] The signal direction finding function has been added to the Bluetooth core specification, providing a new solution for indoor sub-meter real-time positioning. The direction finding function includes the signal arrival angle (Angle of Arrival, AoA) and departure angle (Angle of Departure, AoD), and positioning is performed by measuring the direction of the signal and the spatial installation position of the base station. Compared with Bluetooth RSS positioning, the positioning beacon does not need to maintain a fixed transmission power, so the requirements for positioning tags are lower, the compatibility is stronger, and there are greater advantages in the difficulty and cost of base station deployment. The principle of AoA positioning is as follows Figure 2 shown.
[0005] The positioning accuracy of Bluetooth AoA depends on the accuracy of angle estimation. The angle of arrival of the signal is also one of the basic issues in array signal processing. The Multiple Signal Classification (MUSIC) algorithm is a super-resolution algorithm. When the signals are independent, it has better performance than other types of angle estimation algorithms. It is widely used in radar, sonar and other fields. Some scholars have proposed a dimensionality reduction 2D-MISUC algorithm, which decomposes the 2D direction of arrival (DoA) estimation into two levels of 1D DoA estimation, greatly reducing the amount of calculation, but the accuracy of the angle estimation has not been improved. In addition, when estimating the direction of arrival indoors, due to the complexity of the signal propagation path, there will be coherent sources in the signal incident on the array antenna, such as Figure 3As shown in Figure 1, the antenna array will receive coherent signals from different directions, causing the signal covariance matrix to have a rank deficiency, thereby causing the signal eigenvector to diverge into the noise subspace. At this point, the traditional angle estimation algorithm can no longer effectively estimate the outgoing wave direction.
[0006] Typical decoherence algorithms include spatial smoothing and vector matrix reconstruction. The spatial smoothing algorithm restores the rank of the signal covariance at the expense of the effective aperture of the array antenna. The dimension of the modified matrix is smaller than that of the original matrix, and its performance has certain limitations under low signal-to-noise ratio. The vector matrix reconstruction method uses the received signal to reconstruct the Toeplitz matrix for a two-dimensional decoherence algorithm, but it has requirements on the number and arrangement of array elements. The algorithm can only be applied to cross or L-shaped arrays and the number of antenna elements on each axis must be odd. If the array elements of the array antenna are arranged in a cross or L-shaped manner, the area of the array will inevitably be larger, which will increase certain costs and engineering difficulties in the design and installation of the positioning base station. Summary of the invention
[0007] In order to overcome the deficiencies of the prior art, an object of the present invention is to provide a method and system for measuring the angle of arrival of a Bluetooth signal.
[0008] To achieve the above object, the present invention provides the following solutions:
[0009] A method for measuring the angle of arrival of a Bluetooth signal, comprising:
[0010] Solve the IQ data on different elements of different array antennas to obtain phase information;
[0011] Unwrapping the data of the reference phase in the phase information to obtain an average phase difference;
[0012] Performing phase compensation on the data of each array element according to the average phase difference to obtain a received signal matrix;
[0013] Determine an initial covariance matrix of the array antenna according to the received signal matrix;
[0014] Replacing the main diagonal blocks of the initial covariance matrix with the one-dimensional smoothed covariance matrix of each submatrix in the initial covariance matrix to obtain a first replacement matrix;
[0015] Replacing the upper triangular blocks in the first replacement matrix according to the cross-correlations determined by the pseudo-inverse of the received signals of each sub-matrix in the first replacement matrix to obtain a second replacement matrix;
[0016] Determining a modified covariance matrix using the conjugate symmetry of the second replacement matrix;
[0017] Eigenvalue decomposition and information source number judgment are performed based on the modified covariance matrix, and an angle estimation result is determined according to the two-dimensional spatial spectrum.
[0018] Preferably, the step of calculating IQ data on different array elements of different array antennas to obtain phase information includes:
[0019] The entire array antenna is sampled in turn to obtain the IQ data; each sampling is orthogonal sampling; the IQ data includes an in-phase component and an orthogonal component of the received signal;
[0020] Using formula Determine the phase information; wherein Φ is the phase information, V Q is the orthogonal component, V I is the in-phase component.
[0021] Preferably, the step of unwrapping the data of the reference phase in the phase information to obtain the average phase difference comprises:
[0022] The initial signal matrix is determined according to the phase information; the formula of the initial signal matrix is: Wherein, X is the initial signal matrix, m is the number of sampling times, is the element in the first row and the mth column of the initial signal matrix, Φ i ,i=1,2,3,…,12 is the phase of each antenna array element.
[0023] Using formula Compensate the sampled signal; where φ is the compensation phase;
[0024] The phase sampled in the reference phase is used to perform unwrapping to obtain an average phase difference.
[0025] Preferably, performing phase compensation on the data of each array element according to the average phase difference to obtain a received signal matrix includes:
[0026] Replacing φ in the compensated initial signal matrix with the average phase difference to obtain a compensated phase;
[0027] The compensated phase is mapped to the interval (-π, π] to obtain the received signal matrix.
[0028] Preferably, determining the initial covariance matrix of the array antenna according to the received signal matrix includes:
[0029] Dividing the array antenna to obtain a plurality of sub-arrays;
[0030] The initial covariance matrix is determined according to the number of the sub-arrays and the received signal matrix.
[0031] Preferably, performing eigenvalue decomposition and information source number judgment based on the modified covariance matrix, and determining the angle estimation result according to the two-dimensional spatial spectrum, includes:
[0032] Performing eigenvalue decomposition on the modified covariance matrix to obtain a plurality of eigenvalues and eigenvectors corresponding to the eigenvalues;
[0033] Determine the number of information sources according to the eigenvalue and the eigenvector;
[0034] constructing a spatial spectrum function according to the eigenvector;
[0035] A maximum value search is performed based on the spatial spectrum function to obtain the angle estimation result.
[0036] A system for measuring the angle of arrival of a Bluetooth signal, comprising:
[0037] A calculation unit, used to calculate IQ data on different array elements of different array antennas to obtain phase information;
[0038] A deconvolution unit, configured to deconvolve the data of the reference phase in the phase information to obtain an average phase difference;
[0039] A compensation unit, configured to perform phase compensation on the data of each array element according to the average phase difference to obtain a received signal matrix;
[0040] A covariance determination unit, configured to determine an initial covariance matrix of the array antenna according to the received signal matrix;
[0041] A first dividing unit, used for replacing the main diagonal blocks of the initial covariance matrix according to the one-dimensional smoothed covariance matrix of each sub-matrix in the initial covariance matrix, to obtain a first replacement matrix;
[0042] A second dividing unit is used to replace the upper triangular blocks in the first replacement matrix according to the cross-correlation determined by the pseudo-inverse of the received signal of each sub-matrix in the first replacement matrix to obtain a second replacement matrix;
[0043] A correction unit, used for determining a corrected covariance matrix by using the conjugate symmetry of the second replacement matrix;
[0044] The angle estimation unit is used to perform eigenvalue decomposition and information source number judgment based on the modified covariance matrix, and determine the angle estimation result according to the two-dimensional spatial spectrum.
[0045] Preferably, the solving unit specifically includes:
[0046] A sampling module, used for sampling the entire array antenna in turn to obtain the IQ data; each sampling is orthogonal sampling; the IQ data includes an in-phase component and an orthogonal component of the received signal;
[0047] Phase determination module for using the formula Determine the phase information; wherein Φ is the phase information, V Q is the orthogonal component, V I is the in-phase component.
[0048] Preferably, the unwinding unit specifically comprises:
[0049] The initial matrix determination module is used to determine the initial signal matrix according to the phase information; the formula of the initial signal matrix is: Wherein, X is the initial signal matrix, m is the number of sampling times, is the element in the first row and the mth column of the initial signal matrix, Φ i , i=1,2,3,…,12 is the phase of each antenna element;
[0050] Compensation module for using the formula Compensate the sampled signal; where φ is the compensation phase;
[0051] The unwinding module is used to unwind the phase obtained by sampling in the reference phase to obtain an average phase difference.
[0052] Preferably, the compensation unit specifically includes:
[0053] A compensation module, used to replace φ in the compensated initial signal matrix with the average phase difference to obtain a compensated phase;
[0054] A mapping module is used to map the compensated phase to the interval (-π, π] to obtain the received signal matrix.
[0055] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:
[0056] The present invention provides a method and system for measuring the arrival angle of a Bluetooth signal. The method comprises: solving IQ data on different array elements of different array antennas to obtain phase information; unwinding data of a reference phase in the phase information to obtain an average phase difference; performing phase compensation on data of each array element according to the average phase difference to obtain a received signal matrix; determining an initial covariance matrix of the array antenna according to the received signal matrix; replacing the main diagonal blocks of the initial covariance matrix according to a one-dimensional smoothed covariance matrix of each subarray in the initial covariance matrix to obtain a first replacement matrix; replacing the upper triangular blocks in the first replacement matrix according to the cross-correlation determined by the pseudo-inverse of the received signals of each subarray in the first replacement matrix to obtain a second replacement matrix; determining a modified covariance matrix using the conjugate symmetry of the second replacement matrix; performing eigenvalue decomposition and information source number judgment based on the modified covariance matrix, and determining an angle estimation result according to a two-dimensional spatial spectrum. The present invention corrects the covariance matrix before the spatial spectrum search, which can reduce the arrangement requirements of the antenna array when measuring the arrival angle of the Bluetooth signal, effectively suppress the influence of indoor multipath coherence, and improve the accuracy of signal arrival angle estimation. BRIEF DESCRIPTION OF THE DRAWINGS
[0057] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0058] Figure 1 It is a schematic diagram of the signal strength positioning principle in the prior art;
[0059] Figure 2 It is a schematic diagram of the AoA positioning principle in the prior art;
[0060] Figure 3 A schematic diagram of an antenna array receiving signals in the prior art;
[0061] Figure 4 A flow chart of a measurement method in an embodiment of the present invention;
[0062] Figure 5 A schematic diagram of a receiving antenna receiving a signal in an embodiment of the present invention;
[0063] Figure 6 A schematic diagram of the constant sound expansion package information in the embodiment provided by the present invention;
[0064] Figure 7 This is a first schematic diagram of an antenna array in an embodiment provided by the present invention;
[0065] Figure 8 A second schematic diagram of an antenna array in an embodiment provided by the present invention;
[0066] Fig. 9 A schematic diagram of the first division of the antenna array in the embodiment provided by the present invention;
[0067] Fig.10 A schematic diagram of sub-array overlap in an embodiment of the present invention;
[0068] Fig.11 A schematic diagram of the second division of the antenna array in the embodiment provided by the present invention;
[0069] Fig.12 A schematic diagram of the hardware system design in the embodiment provided by the present invention;
[0070] Fig.13 A flowchart of the algorithm steps for angle estimation in the embodiment provided by the present invention;
[0071] Fig.14 A comparison diagram of pitch angles of changes in the number of snapshots in the embodiment provided by the present invention;
[0072] Fig.15 A comparison diagram of the azimuth angles of the changes in the number of snapshots in the embodiment provided by the present invention;
[0073] Fig.16 A diagram showing the azimuth angle comparison of the signal-to-noise ratio change in the embodiment provided by the present invention;
[0074] Fig.17 A diagram showing the azimuth angle comparison of the signal-to-noise ratio change in the embodiment provided by the present invention;
[0075] Fig.18 A schematic diagram of angle estimation results of a conventional 2D-MUSIC in an embodiment of the present invention;
[0076] Fig.19 A schematic diagram of angle estimation results of the experimental method in the embodiment provided by the present invention;
[0077] Fig. 20 A schematic diagram of angle estimation results of the same static point in conventional 2D-MUSIC in an embodiment provided by the present invention;
[0078] Fig.21 A schematic diagram of angle estimation results of the same static point of the experimental method in the embodiment provided by the present invention;
[0079] Fig. 22 A schematic diagram of angle estimation results of traditional 2D-MUSIC in an indoor environment in an embodiment provided by the present invention;
[0080] Fig.23 A schematic diagram of angle estimation results of the experimental method in an indoor environment in an embodiment provided by the present invention;
[0081] Fig.24 A schematic diagram of two-dimensional mapping of angle information in an embodiment provided by the present invention;
[0082] Fig.25 A schematic diagram of a positioning point angle estimation result of a traditional 2D-MUSIC in an embodiment provided by the present invention;
[0083] Fig.26 This is a schematic diagram of the positioning point angle estimation result of the experimental method in the embodiment provided by the present invention. DETAILED DESCRIPTION
[0084] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0085] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.
[0086] The terms "first", "second", "third" and "fourth" in the specification and claims of the present application and the drawings are used to distinguish different objects rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a series of steps, processes, methods, etc. are not limited to the listed steps, but may optionally include steps that are not listed, or may optionally include other step elements inherent to these processes, methods, products or devices.
[0087] The purpose of the present invention is to provide a method and system for measuring the arrival angle of a Bluetooth signal, which can reduce the arrangement requirements of the antenna array when measuring the arrival angle of a Bluetooth signal, effectively suppress the influence of indoor multipath coherence, and improve the accuracy of signal arrival angle estimation.
[0088] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.
[0089] Figure 4 and Fig.13 They are respectively a flow chart of the measurement method and a flow chart of the algorithm steps of angle estimation in the embodiment provided by the present invention, such as Figure 4 and Fig.13 As shown, the present invention provides a method for measuring the arrival angle of a Bluetooth signal, comprising:
[0090] Step 100: Calculate the IQ data on different array elements of different array antennas to obtain phase information.
[0091] Step 200: unwrap the data of the reference phase in the phase information to obtain an average phase difference.
[0092] Step 300: Perform phase compensation on the data of each array element according to the average phase difference to obtain a received signal matrix.
[0093] Step 400: Determine an initial covariance matrix of the array antenna according to the received signal matrix.
[0094] Step 500: replacing the main diagonal blocks of the initial covariance matrix according to the one-dimensional smoothed covariance matrix of each sub-matrix in the initial covariance matrix to obtain a first replacement matrix.
[0095] Step 600: replace the upper triangular blocks in the first replacement matrix according to the cross-correlations determined by the pseudo-inverse of the received signals of each sub-matrix in the first replacement matrix to obtain a second replacement matrix.
[0096] Step 700: Determine a modified covariance matrix using the conjugate symmetry of the second replacement matrix.
[0097] Step 800: Perform eigenvalue decomposition and information source number determination based on the modified covariance matrix, and determine an angle estimation result according to a two-dimensional spatial spectrum.
[0098] Preferably, the step 100 specifically includes:
[0099] The entire array antenna is sampled in turn to obtain the IQ data; each sampling is orthogonal sampling; the IQ data includes an in-phase component and an orthogonal component of the received signal;
[0100] Using formula Determine the phase information; wherein Φ is the phase information, V Q is the orthogonal component, V I is the in-phase component.
[0101] In this embodiment, Figure 5 As shown, assume that a far-field narrowband signal is incident from the normal direction θ that deviates from the two receiving antennas. The distance between the two receiving antennas A1 and A2 is d, and the antenna distance d is less than 1 / 2 of the signal wavelength λ. Since the wave paths of the signal reaching antennas A1 and A2 are different, the signals received by the two antennas will have a phase difference. Obviously, the phase difference of the signal is related to the deviation angle θ, the wavelength λ of the incident signal, and the distance d between the two antennas, and the relationship can be expressed as formula (1).
[0102]
[0103] If the receiving antennas are spaced at a fixed distance, the signal wavelength is known and remains unchanged during the antenna sampling period, θ can be estimated by the phase difference of the antenna receiving signal. To achieve the direction-finding function, the Bluetooth core specification version 5.1 adds a variable-length Constant Tone Extension (CTE) to the transmitter and receiver data packets, such as Figure 6 This expansion package defines a GFSK modulation wave with a fixed frequency of 16 to 160us and 250kHz, and specifies the time interval for antenna switching and sampling at the receiving end.
[0104] As a low-cost, low-power device, the Bluetooth receiver does not choose to access each antenna array element at the same time as other receivers, but visits them in turn at the same time interval. According to the standard of the Bluetooth core specification, the reference antenna is first used to sample the CTE signal 8 times, each time interval is 1μs. Then, the antenna is switched at a specified time interval during the duration of the CTE signal to sample the entire array in turn. Each sampling is orthogonal sampling, and the in-phase component V of the received signal is obtained. I With the orthogonal component V Q , that is, IQ data. According to the characteristics of orthogonal sampling, the phase information can be obtained through formula (2).
[0105]
[0106] The estimation of the two-dimensional arrival angle of a signal requires a two-dimensional array antenna, which generally adopts an L-shaped, rectangular, cross-shaped, or circular array with uniformly distributed antenna elements. After the array antenna collects data, the direction of the incoming wave can be estimated based on the array model.
[0107] Preferably, the step 200 specifically includes:
[0108] The initial signal matrix is determined according to the phase information; the formula of the initial signal matrix is: Wherein, X is the initial signal matrix, m is the number of sampling times, is the element in the first row and the mth column of the initial signal matrix, Φ i ,i=1,2,3,…,12 is the phase of each antenna array element.
[0109] Using formula The sampled signal is compensated; where φ is the compensation phase.
[0110] The phase sampled in the reference phase is used to perform unwrapping to obtain an average phase difference.
[0111] In this embodiment, the two-dimensional signal arrival angle estimation requires a two-dimensional array antenna, generally using an L-shaped array, rectangular array, cross array, and circular array with uniformly distributed antenna elements. The performance of the array antenna is mainly related to the number of elements, the number of sampled snapshots, and the signal-to-noise ratio. Since the Bluetooth receiver polls the antenna elements and the maximum CTE duration is 160μs, the number of antenna elements and the number of sampled snapshots of the Bluetooth AOA positioning base station are mutually exclusive, and one increases while the other decreases. This embodiment uses a rectangular array of 12 elements, and the distance d between adjacent elements is less than 1 / 2 of the Bluetooth signal wavelength. Figure 7 shown.
[0112] like Figure 8 As shown in the figure, a beam of signal in space is received by the array antenna, and its azimuth and elevation angles are θ, In space, the distance between the source and the receiver It can be regarded as a far-field signal, where D and λ are the array antenna aperture and signal wavelength respectively. Therefore, in Bluetooth AoA applications, it can be regarded as a far-field narrowband signal of the CTE extension package.
[0113] Taking the origin O as the reference point, the distance of each array element from point O and its azimuth are R i ,Θ i ,i=1,2,…,12. Since the wave path of the signal reaching each array element is different, the phase of the signal received by different array elements will be different. The phase difference between the signal received by each array element and the reference point O can be expressed as:
[0114]
[0115] Obviously easy to get:
[0116]
[0117] And due to the geometric position of the array antenna, we have:
[0118]
[0119] The steering vector of the array can be expressed as Then the received signal of the array antenna can be written as
[0120]
[0121] Where s(t) is the transmitted signal, N = [n 1 (t),n 2 (t),…,n 12 (t)] T is the noise vector. If there are n signal sources, the formula becomes: where A is the array manifold, S is the source vector, and N is the noise vector.
[0122]
[0123] The formula represents the ideal data model of the antenna array. The actual data model is usually different due to factors such as physical phenomena in the electromagnetic environment and hardware defects, as shown in the formula.
[0124] X real =OLC·A·MS+N (8)
[0125] In the formula, S represents the signal sent by the real source; M represents the impact of multipath signals on the array model; A is the array manifold, C is the impact of mutual coupling between antenna elements; L is the impact of inconsistency between elements such as the length of the wire path; O is the impact of the frequency error of the local oscillator; N is the independent noise term. The impact of each factor is expressed as a corresponding matrix. The source sends a signal S, and multipath occurs first. After being received by the array, the impact of the array itself on the signal is superimposed. Various influences are generated in sequence, and the order of matrix multiplication in the formula is also determined by this. The OLC term can be weakened by improving the hardware platform and calibrating the array antenna.
[0126] Preferably, the step 300 specifically includes:
[0127] The average phase difference is used to replace φ in the compensated initial signal matrix to obtain the compensated phase.
[0128] The compensated phase is mapped to the interval (-π, π] to obtain the received signal matrix.
[0129] In this embodiment, the Bluetooth receiver switches the antenna switch during the duration of the CTE signal to perform orthogonal sampling on the signal on each antenna array element to obtain the in-phase component V I With the orthogonal component V Q , and the instantaneous phase of each antenna array element is calculated by formula.
[0130] Assuming that each array element samples m times during the duration of the CTE signal, the received signal matrix can be expressed as:
[0131]
[0132] Assume that the time interval between two adjacent samples is τ, and the modulation signal period is T, then each sampling result will have a phase shift with the real signal. The sampling signal phase is compensated as shown in the formula.
[0133]
[0134] In order to reduce the frequency error of the receiver itself, the phase obtained by 8 samplings in the reference phase can be used, and the average phase difference can be obtained after unwrapping. Substituting φ in the formula, the compensated phase is then mapped to the interval (-π,π] to obtain the final received signal matrix
[0135] Preferably, the step 400 specifically includes:
[0136] Dividing the array antenna to obtain a plurality of sub-arrays;
[0137] The initial covariance matrix is determined according to the number of the sub-arrays and the received signal matrix.
[0138] In this embodiment, the array antenna is arranged according to Fig. 9 The four sub-arrays are divided into sub-array 1, sub-array 2, sub-array 3 and sub-array 4, and all of them are uniform linear arrays.
[0139] Fig.10 Each sub-array in can be further divided into two overlapping sub-arrays L1 and L2 with 3 array elements, such as Fig.10 shown.
[0140] Assume that the received signals of L1 and L2 are X L1 ,X L2 , can be obtained Fig.10 The forward smoothed covariance matrix G of neutron array 1 1,1 , as shown in formula (11).
[0141]
[0142] Similarly, we can get Figure 5 The forward smoothing covariance matrix G of the other three sub-matrices in 2,2 ,G 3,3 ,G 4,4 .
[0143] Let J 3 is the third-order anti-unit matrix, as shown in Eq.
[0144]
[0145] The signal covariance matrix and its conjugate inverse matrix are averaged to correct the Toeplitz property of the covariance matrix of the received signal of the one-dimensional uniform linear array, as shown in equation (13), (·) * Represents the conjugate of a matrix.
[0146]
[0147] The array antenna is divided into 4 sub-arrays. Fig.11 As shown, each sub-array has only 3 array elements.
[0148] The received signal of the entire array can be expressed as in(·) T represents the transpose of the matrix, then the received signal The consistent estimate of covariance can be expressed as, where (·) H represents conjugate transpose, and m is the number of snapshots.
[0149]
[0150] From formula (14), we can see that R xx The main diagonal block matrices of are the autocorrelation matrices of the four sub-matrices, and the other blocks are the cross-correlations of the sub-matrices. To replace R in the formula xx The main diagonal block.
[0151] The cross-correlation between subarray 1 and subarray 2 in equation (14) can be expressed as:
[0152]
[0153] In formula (15) are the theoretical received signals without interference noise of subarray 1 and subarray 2, respectively. When the number of signal sources n≥3, and When the number of information sources n<3, and The rank is the number of sources n, due to the noise term N 1 、N 2 The impact of X 1 With X 2 The rank of cannot be determined, but it has a generalized inverse.
[0154] set up For X i The generalized inverse matrix of With X i Satisfies the Moore-Penrose condition:
[0155]
[0156] According to conditions ① and ④ in the formula, we have:
[0157]
[0158] Similarly R 1,2 can be written as:
[0159]
[0160] Formula (17) and formula (18) show that X 1 and X 2 The relationship between the cross-correlation and its autocorrelation. Combining equations and equations, using the corrected autocorrelation in equation By taking the average to reduce the influence of noise, the corrected cross-correlation can be obtained, as shown in formula (19).
[0161]
[0162] Similarly, we can obtain Substituting R in formula (14) xx The block matrices in the upper right corner.
[0163] Because R xx is a Hermitian matrix that satisfies conjugate symmetry, so the corrected covariance matrix of the entire array can be obtained:
[0164]
[0165] Preferably, the step 800 specifically includes:
[0166] Performing eigenvalue decomposition on the modified covariance matrix to obtain a plurality of eigenvalues and eigenvectors corresponding to the eigenvalues;
[0167] Determine the number of information sources according to the eigenvalue and the eigenvector;
[0168] constructing a spatial spectrum function according to the eigenvector;
[0169] A maximum value search is performed based on the spatial spectrum function to obtain the angle estimation result.
[0170] In this embodiment, The formula for eigenvalue decomposition is as follows:
[0171]
[0172] λ in formula (21) 1 ,λ 2 ,…,λ 12 for Since the power of the signal is greater than the power of the noise, the eigenvector U corresponding to the larger n eigenvalues S Then the signal subspace is formed, and the eigenvector U corresponding to the smaller 12-n eigenvalues N Span the noise subspace.
[0173] In general, the number of effective information sources n = 1. In indoor environments, there are often reflected multipath signals. At this time, the number of information sources can be determined based on the number of main eigenvalues of Rxx. Assume that the eigenvalues of Rxx are arranged from large to small, that is, λ 1 ≥λ 2 ≥…≥λ n >>λ n+1 …≥λ 12 , but At this time U N =[ξ n+1 ,ξ n+2 ,…,ξ 12 ].
[0174] In addition, the number of information sources n may be estimated according to Akike Information Theory (AIC) criterion, Minimum Description Length (MDL) criterion, Effective Detection Convergence (EDC) criterion and other methods, but the present invention is not limited thereto.
[0175] byU N The orthogonal relationship with the column vector of the array manifold A can construct the spatial spectrum function, as shown in formula (22).
[0176]
[0177] Finally, through the two-dimensional spatial spectrum search θ∈(-π,π), we can get the direction of the signal.
[0178] The algorithm provided in this embodiment performs a covariance matrix R before searching the spatial spectrum. xx Compared with the MUSIC algorithm, the correction only increases the amount of calculation. When performing angle estimation, the fast MUSIC algorithm can still be used to reduce the amount of calculation. And, the corrected covariance matrix is The dimension is not reduced, so the purpose of weakening the influence of the coherent signal source is achieved without losing the effective aperture of the array antenna. In addition, the algorithm in this article only needs to meet the requirement that the subarray is a uniform linear array with an equal number of array elements, so as long as the division method of the array antenna is planned well, the algorithm provided in this embodiment is still applicable to L-type, cross-type and other arrays.
[0179] In this embodiment, Nordic's nRF52833 chip is used as the Bluetooth receiver. The hardware system block diagram is as follows: Fig.12 As shown in the figure, nRF52833 collects IQ data from different array elements of different array antennas by switching the antenna switch, uploads it to the PC, and then processes the data for angle estimation. The antenna switching time slot and the sampling time slot are both 1μs, that is, the time interval between two adjacent samplings is 2μs.
[0180] In addition, this embodiment also provides a simulation analysis process for this method, which is as follows:
[0181] In order to verify the effectiveness of the algorithm, a Monte Carlo simulation experiment is first designed to analyze the performance of the angle estimation algorithm, and the root mean square error is used to evaluate the performance of the angle estimation, as shown in formula (21).
[0182]
[0183] Where e is the root mean square error, N = 500 represents the number of angle estimates, ψ i represents the angle estimation result of the i-th time, is the actual angle. The signal-to-noise ratio of the array antenna receiving signal is fixed at 2dB, the number of sampling snapshots is changed, and the root mean square error of the elevation angle and azimuth angle is calculated respectively. The results are as follows Fig.14 and Fig.15 shown.
[0184] The number of sampling snapshots is fixed to 6, the signal-to-noise ratio is changed, and the root mean square error of the pitch angle and azimuth angle is calculated respectively. The results are as follows Fig.16 and Fig.17 shown.
[0185] from Fig.14 and Fig.17 It can be seen that in the case of low snapshot and low signal-to-noise ratio, the proposed algorithm has obvious advantages and the error of the calculated angle estimation is smaller.
[0186] The outdoor experiment of this embodiment is carried out in a square with a wide field of vision. The array antenna is installed on a pole 3.8 meters above the ground, and the normal direction of the array is horizontal. The beacon is placed on the ground in front of the base station. Due to the open environment, it can be regarded as having no multipath signal at this time. During the duration of the 160μs CTE signal emitted by the beacon, the receiver can complete a maximum of 6 polling samples of the entire array in addition to 8 samples of the reference antenna. The measured IQ data is solved according to the above solution method to obtain the received signal matrix of the array. Then the angle is estimated using the classic algorithm and the algorithm proposed in this paper. The results are as follows Fig.18 and Fig.19 As shown, θ, Denote the azimuth angle and the elevation angle respectively. The spatial spectrum peak of the algorithm proposed in this embodiment is sharper than that of the classical algorithm and has a better signal-to-noise ratio tolerance, which is consistent with the simulation results.
[0187] Three static points were randomly selected in the distance, and 500 data were collected at each static point. The angle estimation results are shown in Fig. 20 and Fig.21 As shown in the figure, in the outdoor without the influence of coherent sources, the angle estimation results of the two algorithms are basically the same. However, from the same static point, the estimation results of the proposed algorithm are more concentrated.
[0188] Calculate separately Fig. 20 and Fig.21 The root mean square error of the azimuth and elevation angle of each static point in the calculation is shown in Table 1.
[0189] Table 1 Static point comparison diagram
[0190]
[0191] It can be seen from Table 1 that the improved algorithm has a significant improvement in accuracy in estimating the direction of arrival of independent signals.
[0192] The indoor experimental environment is a typical office environment. The array antenna board is fixed on the indoor ceiling, and the normal direction of the array antenna is vertical, with a vertical distance of 3.4 meters from the ground.
[0193] In indoor environments, array antennas may receive multipath signals reflected from walls, desks, floors, and other directions. These co-channel interferences pose a great challenge to the arrival angle estimation of Bluetooth signals. When multipath interference is severe, the classic algorithm is completely ineffective, but the improved algorithm can still work normally. Fig. 22 and Fig.23 As shown, the actual position of the beacon is 180° in azimuth.
[0194] When the tag is located in the normal direction of the array antenna (i.e., 0° elevation angle), the calculated azimuth angle changes greatly, so the angle information is mapped to a two-dimensional plane, such as Fig.24 As shown, the position of the array antenna is the origin O, and the position of the positioning tag is point A. Then the two-dimensional coordinates of the tag are, where H is the vertical height between the horizontal plane where the tag is located and the plane of the array antenna.
[0195] The tag was placed at 5 different positions on the same horizontal plane and valid data was collected 200 times at each position. The azimuth and elevation angles were (0,0), (0,30), (30,30), (90,30), and (270,15) respectively. A total of 1000 data collections were performed on the 5 positioning points, and the calculations were performed using the traditional algorithm and the improved algorithm respectively. The calculation results are shown in the figure below. Fig.25and Fig.26 As shown in the figure, due to the existence of multipath interference, the variance of the results calculated by the traditional algorithm is large and many outliers appear. There is even no obvious boundary between positioning point 1 and positioning point 5. The improved algorithm proposed in this paper has significantly improved this situation.
[0196] Since beacons are located at different locations, they are affected by multipath coherence to different degrees, making it difficult to evaluate the overall situation. Fig.25 and Fig.26 The circular probability error is calculated for each of the five positioning points, as shown in formula (22), where ε represents the circular probability error, x i With y i The coordinate points are calculated according to the angle estimation results each time. and is the real coordinate of the positioning point. The calculation results are shown in Table 2.
[0197]
[0198] Table 2 Comparison of calculation results
[0199]
[0200] It can be seen from Table 2 that the algorithm provided in this embodiment greatly reduces the circular probability error of Bluetooth AoA indoor positioning, and even reduces the error by more than 60% at positioning point 5.
[0201] Corresponding to the above method, this embodiment further provides a system for measuring the angle of arrival of a Bluetooth signal, including:
[0202] A calculation unit, used to calculate IQ data on different array elements of different array antennas to obtain phase information;
[0203] A deconvolution unit, configured to deconvolve the data of the reference phase in the phase information to obtain an average phase difference;
[0204] A compensation unit, configured to perform phase compensation on the data of each array element according to the average phase difference to obtain a received signal matrix;
[0205] A covariance determination unit, configured to determine an initial covariance matrix of the array antenna according to the received signal matrix;
[0206] A first dividing unit, used for replacing the main diagonal blocks of the initial covariance matrix according to the one-dimensional smoothed covariance matrix of each sub-matrix in the initial covariance matrix, to obtain a first replacement matrix;
[0207] A second dividing unit is used to replace the upper triangular blocks in the first replacement matrix according to the cross-correlation determined by the pseudo-inverse of the received signal of each sub-matrix in the first replacement matrix to obtain a second replacement matrix;
[0208] A correction unit, used for determining a corrected covariance matrix by using the conjugate symmetry of the second replacement matrix;
[0209] The angle estimation unit is used to perform eigenvalue decomposition and information source number judgment based on the modified covariance matrix, and determine the angle estimation result according to the two-dimensional spatial spectrum.
[0210] Preferably, the solving unit specifically includes:
[0211] A sampling module, used for sampling the entire array antenna in turn to obtain the IQ data; each sampling is orthogonal sampling; the IQ data includes an in-phase component and an orthogonal component of the received signal;
[0212] Phase determination module for using the formula Determine the phase information; wherein Φ is the phase information, V Q is the orthogonal component, V I is the in-phase component.
[0213] Preferably, the unwinding unit specifically comprises:
[0214] The initial matrix determination module is used to determine the initial signal matrix according to the phase information; the formula of the initial signal matrix is: Wherein, X is the initial signal matrix, m is the number of sampling times, is the element in the first row and the mth column of the initial signal matrix, Φ i , i=1,2,3,…,12 is the phase of each antenna element;
[0215] Compensation module for using the formula Compensate the sampled signal; where φ is the compensation phase;
[0216] The unwinding module is used to unwind the phase obtained by sampling in the reference phase to obtain an average phase difference.
[0217] Preferably, the compensation unit specifically includes:
[0218] A compensation module, used to replace φ in the compensated initial signal matrix with the average phase difference to obtain a compensated phase;
[0219] A mapping module is used to map the compensated phase to the interval (-π, π] to obtain the received signal matrix.
[0220] The beneficial effects of the present invention are as follows:
[0221] The present invention proposes an improved two-dimensional angle estimation algorithm by analyzing the characteristics of Bluetooth signals and array antenna models. When measuring the arrival angle of Bluetooth signals, the arrangement requirements of the antenna array can be reduced, the influence of indoor multipath coherence can be effectively suppressed, and the accuracy of signal arrival angle estimation can be improved.
[0222] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.
[0223] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of the present invention. At the same time, for those skilled in the art, according to the idea of the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.
Claims
1. A method for measuring the angle of arrival of a Bluetooth signal, It is characterized in that include: Solve the IQ data on different elements of different array antennas to obtain phase information; Unwrapping the data of the reference phase in the phase information to obtain an average phase difference; Performing phase compensation on the data of each array element according to the average phase difference to obtain a received signal matrix; Determine an initial covariance matrix of the array antenna according to the received signal matrix; Replacing the main diagonal blocks of the initial covariance matrix with the one-dimensional smoothed covariance matrix of each submatrix in the initial covariance matrix to obtain a first replacement matrix; Replacing the upper triangular blocks in the first replacement matrix according to the cross-correlations determined by the pseudo-inverse of the received signals of each sub-matrix in the first replacement matrix to obtain a second replacement matrix; Determining a modified covariance matrix using the conjugate symmetry of the second replacement matrix; Performing eigenvalue decomposition and information source number judgment based on the modified covariance matrix, and determining an angle estimation result according to a two-dimensional spatial spectrum; The determining the initial covariance matrix of the array antenna according to the received signal matrix specifically includes: Dividing the array antenna to obtain a plurality of sub-arrays; Determining the initial covariance matrix according to the number of the sub-arrays and the received signal matrix; The divided sub-arrays are all uniform linear arrays and the number of array elements in each sub-array is equal.
2. The method for measuring the angle of arrival of a Bluetooth signal according to claim 1, It is characterized in that The step of calculating the IQ data on different array elements of different array antennas to obtain phase information includes: The entire array antenna is sampled in turn to obtain the IQ data; each sampling is orthogonal sampling; the IQ data includes an in-phase component and an orthogonal component of the received signal; Using formula Determine the phase information; wherein Φ is the phase information, V Q is the orthogonal component, V I is the in-phase component.
3. The method for measuring the angle of arrival of a Bluetooth signal according to claim 2, It is characterized in that The step of unwrapping the data of the reference phase in the phase information to obtain an average phase difference comprises: The initial signal matrix is determined according to the phase information; the formula of the initial signal matrix is: Wherein, X is the initial signal matrix, m is the number of sampling times, is the element in the first row and the mth column of the initial signal matrix, Φ i , i=1,2,3,…,12 is the phase of each antenna element; Using formula Compensate the sampled signal; where φ is the compensation phase; The phase sampled in the reference phase is used to perform unwrapping to obtain an average phase difference.
4. The method for measuring the angle of arrival of a Bluetooth signal according to claim 3, It is characterized in that The performing phase compensation on the data of each array element according to the average phase difference to obtain a received signal matrix includes: Replacing φ in the compensated initial signal matrix with the average phase difference to obtain a compensated phase; The compensated phase is mapped to the interval (-π, π] to obtain the received signal matrix.
5. The method for measuring the angle of arrival of a Bluetooth signal according to claim 1, It is characterized in that The performing eigenvalue decomposition and information source number judgment based on the modified covariance matrix, and determining the angle estimation result according to the two-dimensional spatial spectrum, includes: Performing eigenvalue decomposition on the modified covariance matrix to obtain a plurality of eigenvalues and eigenvectors corresponding to the eigenvalues; Determine the number of information sources according to the eigenvalue and the eigenvector; constructing a spatial spectrum function according to the eigenvector; A maximum value search is performed based on the spatial spectrum function to obtain the angle estimation result.
6. A system for measuring the angle of arrival of a Bluetooth signal, It is characterized in that include: A calculation unit, used to calculate IQ data on different array elements of different array antennas to obtain phase information; A deconvolution unit, configured to deconvolve the data of the reference phase in the phase information to obtain an average phase difference; A compensation unit, configured to perform phase compensation on the data of each array element according to the average phase difference to obtain a received signal matrix; A covariance determination unit, configured to determine an initial covariance matrix of the array antenna according to the received signal matrix; A first dividing unit, used for replacing the main diagonal blocks of the initial covariance matrix according to the one-dimensional smoothed covariance matrix of each sub-matrix in the initial covariance matrix, to obtain a first replacement matrix; A second dividing unit is used to replace the upper triangular blocks in the first replacement matrix according to the cross-correlation determined by the pseudo-inverse of the received signal of each sub-matrix in the first replacement matrix to obtain a second replacement matrix; A correction unit, used for determining a corrected covariance matrix by using the conjugate symmetry of the second replacement matrix; An angle estimation unit, used for performing eigenvalue decomposition and information source number judgment based on the modified covariance matrix, and determining an angle estimation result according to a two-dimensional spatial spectrum; The determining the initial covariance matrix of the array antenna according to the received signal matrix specifically includes: Dividing the array antenna to obtain a plurality of sub-arrays; Determining the initial covariance matrix according to the number of the sub-arrays and the received signal matrix; The divided sub-arrays are all uniform linear arrays and the number of array elements in each sub-array is equal.
7. The system for measuring the angle of arrival of a Bluetooth signal according to claim 6, It is characterized in that The solving unit specifically comprises: A sampling module, used for sampling the entire array antenna in turn to obtain the IQ data; each sampling is orthogonal sampling; the IQ data includes an in-phase component and an orthogonal component of the received signal; Phase determination module for using the formula Determine the phase information; wherein Φ is the phase information, V Q is the orthogonal component, V I is the in-phase component.
8. The system for measuring the angle of arrival of a Bluetooth signal according to claim 7, It is characterized in that The unwinding unit specifically comprises: The initial matrix determination module is used to determine the initial signal matrix according to the phase information; the formula of the initial signal matrix is: Wherein, X is the initial signal matrix, m is the number of sampling times, is the element in the first row and the mth column of the initial signal matrix, Φ i , i=1,2,3,…,12 is the phase of each antenna element; Compensation module for using the formula Compensate the sampled signal; where φ is the compensation phase; The unwinding module is used to unwind the phase obtained by sampling in the reference phase to obtain an average phase difference.
9. The system for measuring the angle of arrival of a Bluetooth signal according to claim 8, It is characterized in that The compensation unit specifically includes: A compensation module, used to replace φ in the compensated initial signal matrix with the average phase difference to obtain a compensated phase; A mapping module is used to map the compensated phase to the interval (-π, π] to obtain the received signal matrix.
