A high-precision parameter estimation algorithm based on beamforming BF-SAGE

Abstract

The present invention relates to a beamforming-based BF-SAGE high-precision parameter estimation algorithm, comprising the following steps: step 1) algorithm parameter setting; step 2) using beamforming technology to perform spatial beam scanning; step 3) using the SAGE algorithm to perform parameter estimation on the scanning result; step 4) calculating CRLB to evaluate the angle estimation accuracy of the algorithm; step 5) judging whether the requirements of the communication system are met according to the result of the accuracy evaluation, if the requirements are met, outputting the parameter estimation result, if the requirements are not met, executing step 6); step 6), adjusting the algorithm parameters to the parameters under the non-uniform scanning mode, adjusting the scanning mode to the spatial beam non-uniform scanning, and re-executing step 3)-step 5). Compared with the prior art, the BF-SAGE algorithm of the present invention has higher angle estimation accuracy and higher angle estimation resolution.

Description

Technical Field

[0001] The invention relates to the technical field of wireless communication algorithms, and in particular to a BF-SAGE high-precision parameter estimation algorithm based on beamforming. Background Art

[0002] As 5G technology, standards, and development cycles mature, research hotspots and interests are slowly turning to potential 6G technologies and systems. At present, communication-sensing integration has become a key research direction available in 6G technology. In particular, the research on high-precision positioning plays a vital role in communication-sensing integration technology. High-precision positioning in wireless transmission is usually based on the wave angle of arrival and wave departure angle information of the transmission path with high reliability and high accuracy, so it is particularly important to design a more accurate channel angle estimation algorithm.

[0003] The multiple signal classification (MUSIC) algorithm, Richter's Maximum likelihood estimation (RiMAX) algorithm and SAGE algorithm can effectively estimate the channel characteristics with high accuracy. The algorithm itself also has the characteristics of high resolution and has been applied in various wireless communication systems. However, the high accuracy and high resolution characteristics of the above algorithms for angle estimation are far from meeting the requirements of high-precision positioning research and application. Therefore, it is necessary to further improve the accuracy and resolution of the algorithm estimation.

[0004] At present, BF technology is also widely used in communication transmission measurement and modeling methods. For example, the literature "Y.-H.Cho and J.-J.Kim, "Line-of-sight mimo channel in millimeterwave beamforming system: Modeling and prototype results," in 2015 IEEE 81st Vehicular Technology Conference (VTC Spring), 2015, pp.1–5" introduces a general channel modeling method for beamforming MIMO systems in line-of-sight transmission environments. In the literature "S.-H.Wu, L.-K.Chiu, and J.-W.Wang, "Reconfigurable hybrid beamforming for dual-polarized mmwave mimo channels: Stochastic channel modeling and architectural adaptation methods," IEEE Trans. Commun., vol.66, no.2, pp.741–755, 2018," a dynamic reconfigurable hybrid beamforming system based on channel diversity and antenna array directivity is proposed. They make full use of the high directivity of BF technology to improve the capacity of communication transmission systems and the robustness to interference signals and noise. However, using BF technology in communication systems only improves the directivity of communication and increases the capacity of communication transmission, but does not have the characteristic of estimating the angle of the transmission path. Summary of the invention

[0005] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to propose a BF-SAGE high-precision parameter estimation algorithm, which can be applied in a spatial beam scanning MIMO communication system with BF technology and has higher accuracy and higher angle estimation resolution in angle estimation.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] A BF-SAGE high-precision parameter estimation algorithm based on beamforming includes the following steps:

[0008] Step 1) Algorithm parameter setting;

[0009] Step 2) performing spatial beam scanning using beamforming technology;

[0010] Step 3) using the SAGE algorithm to perform parameter estimation on the scanning results;

[0011] Step 4) Calculate CRLB to evaluate the angle estimation accuracy of the algorithm;

[0012] Step 5) Determine whether the requirements of the communication system are met based on the results of the accuracy assessment.

[0013] If the requirements are met, the parameter estimation results are output.

[0014] If the requirements are not met, proceed to step 6);

[0015] Step 6), adjust the algorithm parameters to the parameters under the non-uniform scanning mode, adjust the scanning mode to spatial beam non-uniform scanning, and re-execute steps 3) to 5).

[0016] The algorithm parameters of step 1) include a beam weight matrix, and the weight matrix determines the number of spatial scans of the beam and the orientation of the beam.

[0017] The general signal model of the spatial beam scanning in step 2) is:

[0018]

[0019] Where T is the measurement period, which is the duration of one channel measurement between all beams generated by the transmitting and receiving ends; P Tx Polarization mode transmission and P Rx The complex attenuation coefficient of the lth path under polarization reception; P Tx / P Rx ∈[1,2], corresponding to two mutually orthogonal linear polarization directions; Ω Tx,l With Ω Rx,l are the departure angle DoD and arrival angle DoA of the lth path; τ l and v l represent time delay and Doppler shift respectively; as well as Refers to P Tx The response of the ath transmitted beam under polarization mode and P Rx The response of the bth receiving beam under polarization mode; ω i,a,b (t) is complex white Gaussian noise.

[0020] When step 2) is performed for the first time, the spatial beam scanning mode is uniform scanning.

[0021] The beam weight matrix expression under the uniform scanning mode is:

[0022]

[0023] Where m = 1, ..., M, n = 1, ..., N, M is the number of elements in the antenna array, N is the number of beams for spatial scanning, and W Rx,m,n is a matrix of dimension M×N.

[0024] The parameter estimation results of step 3) include the time delay τ in the wave arrival and wave departure directions, the Doppler frequency shift v, the horizontal angle φ and the pitch angle θ.

[0025] The calculation formula of step 4) CRLB with respect to the azimuth angle φ is:

[0026]

[0027] Among them, φ is the azimuth angle, M is the number of elements in the antenna array, N is the number of azimuth scanning beams, K is the number of signal transmission frequencies, and γ0 is the signal-to-noise ratio.

[0028] The calculation formula of step 4) CRLB with respect to the pitch angle θ is:

[0029]

[0030] Among them, θ is the elevation angle, M is the number of elements in the antenna array, N is the number of azimuth scanning beams, K is the number of signal transmission frequencies, and γ0 is the signal-to-noise ratio.

[0031] The CRLB calculation formula in step 4) is applied at the receiving end to obtain the azimuth and elevation angles of CRLB in the wave arrival direction, and is applied at the transmitting end to obtain the azimuth and elevation angles of CRLB in the wave departure direction.

[0032] The step 6) adjusts the beam weight matrix in the algorithm parameters to:

[0033]

[0034]

[0035]

[0036]

[0037] Wherein, a=1,…,A; b=1,…,B / 2, A and B respectively represent the number of beams scanned in a certain range of a specific angle Ω containing the path to be estimated and the number of spatial beams scanned in the remaining angle range, Ω1 and Ω2 are the starting angle and the ending angle of the specific angle range; N=A+B; m=1,…,M, M is the number of elements of the antenna array; is an M×N dimensional matrix, and the specific angle Ω is the elevation angle or azimuth angle obtained according to the parameter estimation result.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] (1) The present invention innovatively combines the high directivity of beamforming technology with the high accuracy and high resolution of the SAGE algorithm for angle estimation. Compared with the traditional MUSIC algorithm, SAGE algorithm and RiMAX algorithm, the BF-SAGE algorithm of the present invention has higher accuracy and higher angle estimation resolution in angle estimation.

[0040] (2) The present invention combines uniform scanning and non-uniform scanning, which can not only avoid the increase in measurement time and measurement complexity due to the use of a large number of beams for uniform scanning, but also further improve the accuracy of angle estimation by using non-uniform scanning within a specific angle range.

[0041] (3) The BF-SAGE algorithm of the present invention can well meet the requirements of high-precision positioning, and can be flexibly applied in the 6G system of integrated communication and perception technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a schematic diagram of the process of the present invention;

[0043] Figure 2 Multipath transmission between transmit and receive MIMO antenna arrays using beamforming technology in mobile environments;

[0044] Figure 3 Comparison of RMSEE(φ) obtained by SAGE and BF-SAGE algorithms respectively;

[0045] Figure 4 When φ=90°, the simulated RMSEE(φ) and the theoretically derived Comparison chart;

[0046] Figure 5 are derived from the SAGE and BF-SAGE algorithms respectively. Comparison chart;

[0047] Figure 6 Obtained by SAGE and BF-SAGE algorithms A plot of the relationship between the difference value and the number of scanned beams;

[0048] Figure 7 The maximum RMSEE (φ) and the corresponding 95% confidence interval distribution comparison diagram under two paths obtained by SAGE and BF-SAGE algorithms under different azimuth differences of the two paths;

[0049] Figure 8The BF-SAGE algorithm uses uniform and non-uniform scanning methods to derive the results under different signal-to-noise ratios and wave arrival azimuths. Comparison chart. DETAILED DESCRIPTION

[0050] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0051] like Figure 1 As shown, a BF-SAGE high-precision parameter estimation algorithm based on beamforming includes the following steps:

[0052] Step 1) Algorithm parameter setting;

[0053] Step 2) performing spatial beam scanning using beamforming technology;

[0054] Step 3) using the SAGE algorithm to perform parameter estimation on the scanning results;

[0055] Step 4) Calculate CRLB to evaluate the angle estimation accuracy of the algorithm;

[0056] Step 5) Determine whether the requirements of the communication system are met based on the results of the accuracy assessment.

[0057] If the requirements are met, the parameter estimation results are output.

[0058] If the requirements are not met, proceed to step 6);

[0059] Step 6), adjust the algorithm parameters to the parameters under the non-uniform scanning mode, adjust the scanning mode to spatial beam non-uniform scanning, and re-execute steps 3) to 5).

[0060] Step 1) Algorithm parameter setting.

[0061] The algorithm parameter settings include signal bandwidth, center frequency, number of signal frequencies, antenna element position vectors, and beam weight matrix, wherein the beam weight matrix determines the number of spatial scans of the beam and the orientation of the beam.

[0062] Step 2) Use beamforming technology to perform spatial beam scanning.

[0063] In this embodiment, a spatial beam scanning proposed in step 2) occurs when Figure 2In the communication transmission scenario shown in the figure, the transmitting and receiving ends of the communication system are equipped with MIMO systems composed of M and N antenna array elements respectively. A and B transmitting and receiving beams with different directions are formed at both ends through beamforming technology. These beams realize spatial scanning in the azimuth and elevation angle domains of arrival / departure to receive signals. The communication system described above constructs the communication scenario in this embodiment.

[0064] The general signal model of the spatial beam scanning in step 2) is:

[0065]

[0066]

[0067] Where T is the measurement period, which is the duration of one channel measurement between all beams generated by the transmitting and receiving ends; P Tx Polarization mode transmission and P Rx The complex attenuation coefficient of the lth path under polarization reception; P Tx / P Rx ∈[1,2], corresponding to two mutually orthogonal linear polarization directions; Ω Tx,l With Ω Rx,l are the departure angle DoD and arrival angle DoA of the lth path; τ l and v l represent time delay and Doppler shift respectively; as well as Refers to P Tx The response of the ath transmitted beam under polarization mode and P Rx The response of the bth receiving beam under polarization mode; ω i,a,b (t) is complex white Gaussian noise.

[0068] When step 2) is performed for the first time, the spatial beam scanning mode is uniform scanning.

[0069] The beam weight matrix expression under the uniform scanning mode is:

[0070]

[0071] Where m = 1, ..., W, n = 1, ..., N, M is the number of elements in the antenna array, N is the number of beams for spatial scanning, and W Rx,m,n is a matrix of dimension M×N.

[0072] Step 3) Use the SAGE algorithm to estimate the parameters of the scanning results.

[0073] The parameter estimation results of step 3) include the time delay τ in the wave arrival and wave departure directions, the Doppler frequency shift v, the horizontal angle φ and the pitch angle θ.

[0074] Step 4) Calculate CRLB to evaluate the angle estimation accuracy of the algorithm.

[0075] The calculation formula of step 4) CRLB for the azimuth angle φ is derived as follows:

[0076] Assuming that there is only a single transmission path in the spatial beam scanning parameter signal model, the Fisher Information Matrix (FIM) of φ is calculated by the log-likelihood function Λ(φ; y)

[0077]

[0078] y can be defined as y = vec[y n (f); f=f1,…,f k ,n=1,…,N],vec[·] represents vector operation,

[0079]

[0080]

[0081]

[0082] Will and Substituting into the Λ(φ; y) formula, we can get FIMF φφ The expression of (Θ) is:

[0083]

[0084] Will Apply to the above formula F φφ (Θ) expression, we get the expression of CRLB(φ):

[0085]

[0086] Among them, φ is the azimuth angle, M is the number of elements in the antenna array, N is the number of azimuth scanning beams, K is the number of signal transmission frequencies, and γ0 is the signal-to-noise ratio.

[0087] Similarly, the calculation formula of CRLB for the pitch angle θ can be derived as follows:

[0088]

[0089] Among them, θ is the elevation angle, M is the number of elements in the antenna array, N is the number of azimuth scanning beams, K is the number of signal transmission frequencies, and γ0 is the signal-to-noise ratio.

[0090] The CRLB calculation formula is applied at the receiving end to obtain the azimuth and elevation angles of CRLB in the wave arrival direction, and is applied at the transmitting end to obtain the azimuth and elevation angles of CRLB in the wave departure direction.

[0091] Step 5) Determine whether the requirements of the communication system are met based on the results of the accuracy evaluation. If the requirements are met, output the parameter estimation results. If the requirements are not met, execute step 6).

[0092] Step 6), adjust the algorithm parameters to the parameters under the non-uniform scanning mode, adjust the scanning mode to spatial beam non-uniform scanning, and re-execute steps 3)-5)

[0093] In step 6), based on the results obtained by uniform scanning estimation, especially the estimated angle parameters, non-uniform scanning estimation is performed, focusing on performing more intensive scanning within the angle range on both sides of the angle result obtained by uniform scanning estimation, and estimating and obtaining more accurate path angle information.

[0094] In order to make the BF-SAGE algorithm more intelligent in angle estimation, so as to achieve more accurate positioning and tracking in a certain direction according to the requirements of the communication system itself, the general expression of the beam weight matrix of the BF-SAGE algorithm is obtained by changing the weight distribution ratio to obtain the beamforming weight under non-uniform scanning mode.

[0095]

[0096]

[0097]

[0098]

[0099] Wherein, a=1,…,A; b=1,…,B / 2, A and B respectively represent the number of beams scanned in a certain range of a specific angle Ω containing the path to be estimated and the number of spatial beams scanned in the remaining angle range, Ω1 and Ω2 are the starting angle and the ending angle of the specific angle range; N=A+B; m=1,…,M, M is the number of elements of the antenna array; is an M×N dimensional matrix, and the specific angle Ω is the elevation angle or azimuth angle obtained according to the parameter estimation result.

[0100] This embodiment verifies the accuracy of the BF-SAGE algorithm applied to the azimuth of arrival estimation at the receiving end.

[0101] In order to simplify the verification of estimation accuracy, the transmission path is set to one, and the true values ​​of the transmission delay and horizontal azimuth angle of arrival of this path are 31.556ns and 85.789474° respectively. On the receiving uniform linear array with 5 antenna elements, the BF-SAGE algorithm and the traditional SAGE algorithm were used to perform 500 Monte Carlo simulations to estimate the delay and horizontal azimuth angle of arrival under different signal-to-noise ratios. When using the BF-SAGE algorithm for estimation, the BF technology is first used to form 10 directional beams to achieve uniform spatial scanning, and then the directional beams with the same number of beams are used to perform non-uniform scanning estimation in space. Focusing on the angle domain where the half-power beam width of the uniform linear array is added to the left and right sides of the angle result obtained by the uniform scanning estimation, 8 beams are used for dense scanning, and the remaining 2 beams are used for sparse scanning in the remaining angle range. There is no significant difference in the delay between the BF-SAGE algorithm and the traditional SAGE algorithm. The delay estimated by the BF-SAGE algorithm is 31.556145ns, while the delay estimated by the traditional SAGE algorithm is 31.5562ns, indicating that the BF-SAGE algorithm does not have a significant improvement in the accuracy of delay estimation. However, the BF-SAGE algorithm has a significant improvement in the accuracy of the azimuth of arrival estimation. Figure 3 It can be clearly seen that when the signal-to-noise ratio is equal to 5dB, the root mean square error of the BF-SAGE algorithm in estimating the azimuth of arrival is -30dB, which is nearly an order of magnitude higher than the root mean square error estimated by the traditional SAGE algorithm. In addition, as the signal-to-noise ratio increases, the accuracy of the traditional SAGE algorithm in estimating the azimuth of arrival will no longer increase and tend to be stable, while the accuracy of the BF-SAGE algorithm in estimating the azimuth of arrival will also increase with the increase of the signal-to-noise ratio.

[0102] The RMSEE(φ) obtained by using the BF-SAGE algorithm to estimate the azimuth of arrival through 500 Monte Carlo simulations is compared with the derived The angle estimation accuracy of the BF-SAGE algorithm is evaluated by comparative analysis. Figure 4 As shown in the figure, the RMSEE(φ) obtained by simulation is consistent with the derived Very close, their values ​​will decrease with the increase of the number of antenna elements and the number of spatial scanning beams. This shows that the accuracy of angle estimation by the BF-SAGE algorithm will increase with the increase of the number of antenna elements and the number of spatial scanning beams. Figure 4 It can also be seen that compared with increasing the number of beams, increasing the number of antenna elements has a greater impact on improving the angle estimation accuracy of the BF-SAGE algorithm.

[0103] In order to further evaluate the performance of the proposed BF-SAGE algorithm for angle estimation accuracy, in step 4), the traditional SAGE algorithm and the BF-SAGE algorithm were compared for angle estimation using antenna arrays of the same size and under the premise of the same parameter settings. Figure 5 It can be clearly seen that when the BF-SAGE algorithm uses 30 directional beams to scan the received signal in the azimuth of arrival, the Compared with the SAGE algorithm This indicates that the angle estimation using the BF-SAGE algorithm has a higher angle estimation accuracy, which further proves that the BF-SAGE algorithm has better performance in spatial positioning and spatial tracking applications.

[0104] In addition, the present invention also analyzes the results derived from the BF-SAGE algorithm and the SAGE algorithm as the spatial scanning beam increases. The relationship between them. Figure 6 It can be seen that when the number of spatial scanning beams is less than 30, the two algorithms derive The difference between the two increases sharply, and then the incremental speed of the difference begins to decrease gradually. When the number of beams in the spatial scan approaches 50, the difference becomes very small, and as the number of beams increases again, The difference between them also tends to be smaller and smaller. This means that when the antenna array uses the BF-SAGE algorithm instead of the SAGE algorithm for angle estimation, the estimation accuracy will not be increased indefinitely with the increase of the spatial scanning beam. When the number of spatial scanning beams reaches a certain value, the increase in the angle estimation accuracy of the BF-SAGE algorithm will become very small. As the number of beams continues to increase, the angle estimation accuracy will not be improved significantly, but its growth will become smaller and smaller. In other words, when the BF-SAGE algorithm is used on the antenna array, there will be an optimal number of spatial scanning beams to estimate the angle, which can not only ensure the maximum improvement of the angle estimation accuracy, but also avoid the increase in the measurement complexity and the increase in the cost of the measurement equipment caused by the increase in the number of unnecessary spatial scanning beams. In addition, the present invention also analyzes and evaluates the performance of the BF-SAGE algorithm in angle estimation resolution. The dual-path spatial beam scanning parameter signal model is considered to be applied to the simulation of this evaluation, and it is also assumed that the BF-SAGE algorithm uses 30 directional beams to scan the received signal in the wave arrival azimuth. In the case of different azimuth angle differences between two transmission paths, the maximum RMSEE(φ) and the corresponding 95% confidence interval are taken from the RMSEE(φ) of the two groups of paths estimated by the BF-SAGE and SAGE algorithms using 500 Monte Carlo simulations to estimate the azimuth angle of arrival. Figure 7It can be clearly found from the distribution comparison that after traversing all the cases of arrival azimuth angle difference, the maximum RMSEE (φ) in the dual path estimated by the BF-SAGE algorithm is smaller than that estimated by the SAGE algorithm. At the same time, in all cases of arrival azimuth angle difference, the maximum value of the 95% confidence interval estimated by the BF-SAGE algorithm is much smaller than the minimum value of the 95% confidence interval estimated by the SAGE algorithm. In addition, when the arrival azimuth angle difference of the two transmission paths is greater than 25°, the maximum RMSEE (φ) in the dual path estimated by the BF-SAGE and SAGE algorithms are relatively small. When the arrival azimuth angle difference continues to increase, the maximum RMSEE (φ) under the SAGE algorithm begins to increase sharply, while the maximum RMSEE (φ) estimated by the BF-SAGE algorithm simulation is still increasing slowly. It is not until the path arrival azimuth angle difference reaches about 3° that the maximum RMSEE (φ) estimated by the BF-SAGE algorithm begins to increase significantly. These observations clearly show that the BF-SAGE algorithm improves the angle resolution of the AoA estimation by nearly 20° compared to the traditional SAGE algorithm. When the same array of 5 antenna elements is used for angle estimation, the SAGE algorithm cannot distinguish the two transmission paths when the angle difference is less than 25°, while the BF-SAGE algorithm can separate them and accurately estimate the angle of each path.

[0105] By analyzing and comparing the two groups of CRLB(φ) obtained under uniform and non-uniform scanning modes, the performance of angle estimation of the proposed BF-SAGE algorithm non-uniform scanning mode within a specific angle range is evaluated.

[0106] In the simulation, A=80, B=20, φ1=70°, φ2=100° are set, and other parameters are set the same as the uniform scanning method, and we can get Figure 8 The result from Figure 8 , the BF-SAGE algorithm uses uniform and non-uniform scanning methods to derive the By comparison, it can be seen that in the range of 70° to 100° of the wave arrival azimuth, the non-uniform scanning method of the BF-SAGE algorithm obtains This proves that the BF-SAGE algorithm using non-uniform scanning can indeed improve the accuracy of angle estimation within a certain angle range.

[0107] The preferred specific embodiments of the present invention are described in detail above. It should be understood that a person skilled in the art can make many modifications and changes based on the concept of the present invention without creative work. Therefore, any technical solution that can be obtained by a person skilled in the art through logical analysis, reasoning or limited experiments based on the concept of the present invention on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A beamforming-based BF-SAGE high-precision parameter estimation algorithm, characterized in that: The following steps are involved: Step 1) Algorithm parameter setting; Step 2) performing spatial beam scanning using beamforming technology; Step 3) using the SAGE algorithm to perform parameter estimation on the scanning results; Step 4) Calculate CRLB to evaluate the angle estimation accuracy of the algorithm; Step 5) Determine whether the requirements of the communication system are met based on the results of the accuracy assessment. If the requirements are met, the parameter estimation results are output. If the requirements are not met, proceed to step 6); Step 6), adjust the algorithm parameters to the parameters under the non-uniform scanning mode, adjust the scanning mode to spatial beam non-uniform scanning, and re-execute steps 3) to 5); The algorithm parameters of step 1) include a beam weight matrix, and the weight matrix determines the number of spatial scans of the beam and the orientation of the beam; The general signal model of the spatial beam scanning in step 2) is: Where T is the measurement period, which is the duration of one channel measurement between all beams generated by the transmitting and receiving ends; P Tx Polarization mode transmission and P Rx The complex attenuation coefficient of the lth path under polarization reception; P Tx / P Rx ∈[1,2], corresponding to two mutually orthogonal linear polarization directions; Ω Tx,l With Ω Rx,l are the departure angle DoD and arrival angle DoA of the lth path; τ l and v l represent time delay and Doppler shift respectively; as well as Refers to P Tx The response of the ath transmitted beam under polarization mode and P Rx The response of the bth receiving beam under polarization mode; ω i,a,b (t) is complex Gaussian white noise; When step 2) is first performed, the spatial beam scanning mode is uniform scanning; The beam weight matrix expression under the uniform scanning mode is: Where m = 1, ..., M, n = 1, ..., N, M is the number of elements in the antenna array, N is the number of beams for spatial scanning, and W Rx,m,n is a matrix of dimension M×N; The step 6) adjusts the beam weight matrix in the algorithm parameters to: Wherein, a=1,…,A; b=1,…,B / 2, A and B respectively represent the number of beams scanned in a certain range of a specific angle Ω containing the path to be estimated and the number of spatial beams scanned in the remaining angle range, Ω1 and Ω2 are the starting angle and the ending angle of the specific angle range; N=A+B; m=1,…,M, M is the number of elements of the antenna array; is an M×N dimensional matrix, and the specific angle Ω is the elevation angle or azimuth angle obtained according to the parameter estimation result.

2. The BF-SAGE high-precision parameter estimation algorithm based on beamforming according to claim 1, characterized in that: The parameter estimation results of step 3) include the time delay τ in the wave arrival and wave departure directions, the Doppler frequency shift v, the horizontal angle φ and the pitch angle θ.

3. The BF-SAGE high-precision parameter estimation algorithm based on beamforming according to claim 1, characterized in that: The calculation formula of step 4) CRLB with respect to the azimuth angle φ is: Among them, φ is the azimuth angle, M is the number of elements in the antenna array, N is the number of azimuth scanning beams, K is the number of signal transmission frequencies, and γ0 is the signal-to-noise ratio.

4. The BF-SAGE high-precision parameter estimation algorithm based on beamforming according to claim 1, characterized in that: The calculation formula of step 4) CRLB with respect to the pitch angle θ is: Among them, θ is the elevation angle, M is the number of elements of the antenna array, N is the number of elevation scanning beams, K is the number of signal transmission frequencies, and γ0 is the signal-to-noise ratio.

5. The BF-SAGE high-precision parameter estimation algorithm based on beamforming according to claim 3 or 4, characterized in that: The CRLB calculation formula in step 4) is applied at the receiving end to obtain the azimuth and elevation angles of CRLB in the wave arrival direction, and is applied at the transmitting end to obtain the azimuth and elevation angles of CRLB in the wave departure direction.

Images