Method for height measurement based on multiple signal classification of meter wave multiple input multiple output radar

Abstract

The application provides a multi-signal classification height measurement method based on a meter wave multiple-input multiple-output radar, and belongs to the technical field of radar signal processing. The method comprises the following steps: modeling a received signal, and calculating an output signal; calculating array data covariance, obtaining a noise subspace, using a generalized multi-signal classification method to obtain an initial value of a direct wave, and using multi-signal classification to obtain an initial value of a horizontal polarization wave reflection coefficient; alternately searching in the order of estimating a direct wave estimated angle initial value and estimating a horizontal polarization wave reflection estimated coefficient initial value, and obtaining an estimated parameter group; judging the estimated parameter group, if a judgment condition is met, outputting a direct wave estimated angle, otherwise returning to alternately search; using the direct wave estimated angle to calculate a target height, and completing multi-signal classification height measurement. The application solves the problems of the existing guide vector synthesis algorithm generating redundant spectral peaks and the problem of sharp increase in operation amount caused by dense grid division search.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of radar signal processing, and particularly relates to a multiple signal classification height measurement method based on a meter wave multiple input multiple output radar. BACKGROUND

[0002] Early meter wave radars are widely used in the field of long-range target early warning due to their small propagation attenuation, long detection distance and simple implementation, but the long wavelength of the meter wave radar also leads to a decrease in resolution, so that the meter wave radar cannot be applied to tracking guidance and other tasks. In modern warfare, emerging stealth technology directly drives the development of anti-stealth technology, and the meter wave radar has natural advantages in anti-stealth and anti-anti-radiation missile aspects, and has once again become a research hotspot. However, when detecting low-altitude targets, the meter wave radar has a very wide beam due to the long electromagnetic wavelength, which leads to serious multipath effects and further causes problems such as lobe splitting. The coherence of the multipath signal and the direct wave signal makes some subspace direction of arrival (DOA) estimation algorithms unusable. The low-altitude height measurement (equivalent to DOA estimation) of the meter wave radar has become a problem to be solved in the field of radar.

[0003] Multiple input multiple output (MIMO) radar is a new system radar widely studied in recent years, unlike the traditional array radar, its unique waveform diversity capability makes it have the advantages of strong anti-interception and anti-jamming capability and virtual array elements. The application of the MIMO system to the meter wave radar can make the receiving virtual beam narrower, which can effectively alleviate the influence of the multipath effect. However, in the meter wave MIMO radar, due to the influence of the transmitted orthogonal waveform, the phenomenon of mutual penetration between the virtual steering vectors occurs, which leads to the fact that the traditional decorrelation algorithm cannot be directly applied, and further makes the MUSIC algorithm unable to be applied to the height measurement of the meter wave MIMO radar. The prior art derives the application of the generalized MUSIC algorithm in the meter wave MIMO radar, and the algorithm does not need to perform decorrelation processing. The prior art reduces the calculation amount by using the dimension reduction method, but needs certain prior knowledge. SUMMARY

[0004] In view of the above problems in the prior art, the multiple signal classification height measurement method based on the meter wave multiple input multiple output radar provided by the application solves the problems of the generation of redundant spectral peaks by the existing steering vector synthesis algorithm and the sharp increase in the amount of calculation caused by the dense grid division search.

[0005] In order to achieve the above purposes, the technical scheme adopted by the application is as follows: a multiple signal classification height measurement method based on a meter wave multiple input multiple output radar, comprising the following steps:

[0006] S1, according to the antenna array and the height measurement target, the millimeter wave multiple-input multiple-output radar multipath receiving signal is modeled, and the received multiple-input multiple-output array receiving signal is matched filtered and vectorized, and the output signal is calculated;

[0007] S2, according to the output signal, the array data covariance is calculated and combined into a covariance matrix, the noise subspace is obtained according to the covariance matrix, the initial value of the direct wave is obtained by using the generalized multiple signal classification method, and the initial value of the horizontal polarization wave reflection coefficient is obtained by using the multiple signal classification;

[0008] S3, according to the generalized multiple signal classification method and the multiple signal classification method, the estimated angle initial value of the direct wave and the estimated coefficient initial value of the horizontal polarization wave reflection are alternately searched in the order of S2, and an estimated parameter group containing the estimated angle of the direct wave and the estimated coefficient of the horizontal polarization wave reflection is obtained;

[0009] S4, define the precision threshold and the iteration threshold, and judge the estimated parameter group, if the estimated parameter group precision is less than the precision threshold or the estimated parameter group iteration number is greater than the iteration threshold, the direct wave estimated angle is output, otherwise return to step S3;

[0010] S5, the direct wave estimated angle is used, and the projection distance between the antenna array and the height measurement target in the millimeter wave multiple-input multiple-output radar multipath receiving signal model and the antenna array height are combined to calculate the target height, and the multiple signal classification height measurement is completed.

[0011] The beneficial effects of the present application are: the present application models the millimeter wave multiple-input multiple-output radar multipath receiving signal, obtains a relatively ideal signal model, and alternately searches the generalized multiple signal classification and the multiple signal classification, avoids the sharp increase of operation amount caused by dense grid division search, realizes good low elevation height measurement effect, and makes the method of the present application not need to solve coherence, and has high engineering practical value.

[0012] Further, the S1 comprises the following steps:

[0013] S101, according to the antenna array and the height measurement target, the millimeter wave multiple-input multiple-output radar multipath receiving signal is modeled;

[0014] S102, according to the millimeter wave multiple-input multiple-output radar multipath receiving signal model, the transmitting signal steering vector and the multipath steering vector of the transmitting signal, the signal of the height measurement target irradiated by the radar is obtained;

[0015] S103, according to the signal of the height measurement target irradiated by the radar and the array element of the antenna array, the echo signal corresponding to the array element is obtained, and the multiple-input multiple-output array receiving signal is obtained by using the echo signal of each array element;

[0016] S104, perform matched filtering and vectorization on the multiple-input multiple-output array received signal to obtain an output signal.

[0017] The above further scheme has the beneficial effect that: the present application models the rice wave multiple-input multiple-output radar multipath received signal, improves the accuracy when using the corresponding data in the model, and accurately receives the echo signal through the elements in the antenna array, obtains the output signal by performing matched filtering and vectorization on the multiple-input multiple-output array received signal, and realizes that the noise of the matched filter output is Gaussian white noise under the consideration of noise.

[0018] Further, the S2 comprises the following steps:

[0019] S201, calculate array data covariance according to the output signal, combine the array data covariance into a covariance matrix, and perform eigenvalue decomposition on the covariance matrix to obtain a noise subspace;

[0020] S202, obtain an initial value of a direct wave estimated angle using a generalized multiple signal classification method according to the noise subspace;

[0021] S203, calculate a steering vector of received data using a horizontal polarization wave reflection coefficient and a direct wave angle according to the rice wave multiple-input multiple-output radar multipath received signal model, and obtain an initial value of a horizontal polarization wave reflection estimated coefficient using the steering vector of the received data and the noise subspace according to the multiple signal classification.

[0022] Further, the expression of the multiple signal classification is as follows:

[0023] ;

[0024] ;

[0025] ;

[0026] wherein, represents a search space spectrum value of the direct wave angle and the reflection coefficient, represents a steering vector of received data, represents a direct wave angle, represents a horizontal polarization wave reflection coefficient, represents a steering vector of a transmitted signal, represents a multipath steering vector of a transmitted signal, represents a reflection angle, represents a phase difference caused by a path difference of a direct wave and a reflected wave, represents a vector product, represents a steering vector of an echo signal, represents an antenna array height, This indicates the projected distance between the antenna array and the target being measured on the horizontal plane.

[0027] The beneficial effects of the above-mentioned further solutions are as follows: the present invention eliminates the influence of multipath effects by using generalized multiple signal classification that does not depend on the reflection coefficient of horizontal polarization waves; the present invention solves the problem of mutual penetration between steering vectors by converting the multipath influence into a signal synthesized by steering vectors and using multiple signal classification that includes the reflection coefficient of horizontal polarization waves, thereby improving the orthogonality between the noise subspace and the steering vectors and ensuring the accuracy and reliability of angle estimation.

[0028] Furthermore, in S3, the initial values ​​of the direct wave estimation angle and the initial values ​​of the horizontal polarized wave reflection estimation coefficients are searched alternately. The specific steps of the alternate search are as follows:

[0029] In the first search, the reflection coefficient of the horizontally polarized wave is estimated using the random direct wave angle as the initial value, and the initial estimated value of the emission coefficient of the horizontally polarized wave is obtained.

[0030] The initial estimate of the horizontally polarized wave emission coefficient is used as the initial value to estimate the angle of the direct wave, thus obtaining the initial estimate of the angle of the direct wave.

[0031] Except for the first search, the reflection coefficient of the horizontally polarized wave is estimated using the current estimate of the direct wave angle, and the next estimate of the emission coefficient of the horizontally polarized wave is obtained.

[0032] The angle of the direct wave is estimated by using the next estimate of the horizontal polarization wave emission coefficient, thus obtaining the next estimate of the angle of the direct wave and realizing alternating search.

[0033] The beneficial effects of the above-mentioned further scheme are as follows: By using alternating search, the present invention estimates the direct wave angle and the reflection coefficient of the horizontally polarized wave, reduces the computational load of multidimensional multi-signal classification and shrinkage, effectively combines the generalized multi-signal classification algorithm and the multi-signal classification algorithm, improves the accuracy of the elevation angle and direction of arrival, and reduces the computational complexity.

[0034] Furthermore, S4 is specifically as follows:

[0035] Define the precision threshold using the iteration precision, and define the iteration threshold using the maximum number of iterations;

[0036] The estimated parameter set is judged by the accuracy threshold and the iteration threshold. If the accuracy of the estimated parameter set is less than the accuracy threshold or the number of iterations of the estimated parameter set is greater than the iteration threshold, the estimated angle is output; otherwise, return to step S3.

[0037] Furthermore, the expression for calculating the target height is as follows:

[0038] ;

[0039] wherein, represents a target height, represents a projection distance between the antenna array and the height measurement target, represents an estimated angle, represents an antenna array height.

[0040] The above further scheme has the beneficial effect that the present application selects the correct spectrum peak in the spatial spectrum function and eliminates false peaks, avoids the influence of redundant false spectrum peaks, improves the accuracy of the estimation method, and makes the alternative search avoid falling into local optimum. BRIEF DESCRIPTION OF DRAWINGS

[0041] Figure 1 The method flowchart of the present application.

[0042] Figure 2 The model diagram obtained by modeling the multi-input multi-output radar multipath receiving signal of the meter wave in the present embodiment.

[0043] Figure 3 The redundant spectrum peak phenomenon diagram in the present embodiment.

[0044] Figure 4 The absolute value of the sum of the signal steering vector and the multipath steering vector at different wavelengths and antenna heights and the direct wave angle relationship diagram in the present embodiment.

[0045] Figure 5 The angle measurement mean square error diagram in the present embodiment.

[0046] Figure 6 The height measurement mean square error diagram in the present embodiment. DETAILED DESCRIPTION

[0047] The specific embodiments of the present application are described below to facilitate the understanding of the present application by those skilled in the art, but it should be clear that the present application is not limited to the scope of the specific embodiments, and for those skilled in the art, any changes that are obvious within the spirit and scope of the present application as defined and determined by the appended claims are obvious, and all inventions utilizing the concept of the present application are within the scope of protection.

[0048] Before the present embodiment is described, the following terms are explained:

[0049] MIMO: multiple-input multiple-output;

[0050] MUSIC: multiple signal classification;

[0051] DOA: direction of arrival;

[0052] SNR: signal-to-noise ratio.

[0053] Embodiment

[0054] In this embodiment, in order to meet the requirements of engineering implementation, the application proposes an alternating search method without solving coherence on the basis of the generalized MUSIC algorithm and the MUSIC algorithm, ensures high accuracy of the pitch angle DOA through iteration, and simplifies the spectrum peak search of the two algorithms to one-dimensional search, greatly reducing the computational complexity.

[0055] As shown in the figure, the application provides a multiple signal classification height measurement method based on a meter wave multiple input multiple output radar, and the implementation method is as follows: Figure 1

[0056] S1, according to the antenna array and the height measurement target, model the meter wave multiple input multiple output radar multipath receiving signal, and perform matched filtering and vectorization on the received multiple input multiple output array receiving signal, and calculate the output signal, the specific steps are as follows:

[0057] S101, according to the antenna array and the height measurement target, model the meter wave multiple input multiple output radar multipath receiving signal, and obtain the meter wave multiple input multiple output radar multipath receiving signal model;

[0058] S102, according to the transmission signal steering vector and the multipath steering vector of the transmission signal in the meter wave multiple input multiple output radar multipath receiving signal model, obtain the signal of the height measurement target irradiated by the radar;

[0059] S103, according to the signal of the height measurement target irradiated by the radar and the array element of the antenna array, obtain the echo signal corresponding to the array element, and obtain the multiple input multiple output array receiving signal using the echo signal of each array element;

[0060] S104, perform matched filtering and vectorization on the multiple input multiple output array receiving signal, and calculate the output signal.

[0061] In this embodiment, as shown in the figure, a transceiver co-located MIMO radar system with Figure 2 is considered M , the array is a uniform linear array in the vertical direction, without loss of generality, it is assumed that the electromagnetic wave polarization mode is horizontal polarization (the electromagnetic wave polarization mode will affect the calculation of the reflection coefficient, which can be set according to the actual situation). It is assumed that the transmission signal vector is: , represents a complex domain value with a dimension of , and has a constraint expression as shown below:

[0062] ;

[0063] ​in, express M An identity matrix of order 1. Indicates the pulse width. Represents the transmitted signal vector;

[0064] exist Figure 2 middle, Indicates the angle of the direct wave. Indicates the reflection angle. This indicates the distance between the antenna array and the target being measured. The distance between the antenna array and the target on the horizontal plane is represented by ; and the multipath signal path length of the distance between the antenna array and the image of the target is denoted as . ;

[0065] In this embodiment, the antenna array is a monostatic array with both transmit and receive signals. Therefore, the transmit signal steering vector and the echo signal steering vector of the antenna array are equal, and the multipath steering vector of the transmit signal is equal to the echo signal steering vector. The signal of the height-measuring target illuminated by the radar can be expressed as:

[0066] ;

[0067] ;

[0068] ;

[0069] ;

[0070] in, This indicates that the target being measured is illuminated by radar. Indicates the steering vector of the transmitted signal. This represents the multipath steering vector of the transmitted signal. Multipath steering vector of echo signal, Indicates the element spacing. This represents the phase difference caused by the path difference between the direct wave and the reflected wave. Indicates the wavelength of electromagnetic waves. This represents the horizontal polarization wave reflection coefficient related to the ground environment;

[0071] Get the first m The expression for the echo signal received by each array element is as follows:

[0072] ;

[0073] in, Indicates the first m The echo signal received by each array element The unknown definite complex constant represents the reflection coefficient between different pulses of a single target. represents the noise received by the m-th array element; and the MIMO array received data is obtained from the echo signal received by the array element, and the expression is as follows: m

[0074]

[0075] wherein, represents the MIMO array received data, and ; and the output signal is obtained by performing matched filtering and vectorization on the MIMO array received data, and the expression of the output signal is as follows:

[0076]

[0077] wherein, represents the output signal, and , represents the vector product, represents the matrix vectorization.

[0078] In this embodiment, considering the noise, it is assumed that the noise received by the array edge is zero-mean Gaussian random white noise, and the expression of the matched filter output result is as follows:

[0079]

[0080] wherein, represents the noise of the matched filter output, represents the original noise vector; the output noise is still Gaussian white noise.

[0081] S2, array data covariance is calculated according to the output signal and combined into a covariance matrix, the noise subspace is obtained according to the covariance matrix, the initial value of the direct wave is obtained by using the generalized multiple signal classification method, and the initial value of the horizontal polarization wave reflection coefficient is obtained by using the multiple signal classification, and the specific steps are as follows:

[0082] S201, array data covariance is calculated according to the output signal, the array data covariance is combined into a covariance matrix, and the covariance matrix is subjected to eigenvalue decomposition to obtain the noise subspace;

[0083] S202, according to the noise subspace, the initial value of the direct wave estimation angle is obtained by using the generalized multiple signal classification method;

[0084] S203, according to the millimeter wave multiple input multiple output radar multipath receiving signal model, the horizontal polarization wave reflection coefficient and the direct wave angle are used to calculate the steering vector of the received data, and according to the multiple signal classification, the initial value of the horizontal polarization wave reflection estimation coefficient is obtained by using the steering vector of the received data and the noise subspace. ​​​​

[0085] In this embodiment, the expression of the array data covariance matrix is as follows:

[0086] ;

[0087] wherein, represents the array data covariance matrix, represents the output signal, represents the number of samples;

[0088] and the noise subspace is obtained by performing eigenvalue decomposition on the array data covariance matrix ;

[0089] According to the noise subspace , the initial value of the direct wave estimated angle is calculated by using the generalized MUSIC which is independent of the horizontal polarization wave reflection coefficient , and the spatial spectrum function of the generalized MUSIC is as follows:

[0090] ;

[0091] wherein, represents the determinant, represents the steering matrix, and the angle of arrival corresponding to the signal source is obtained at the peak of the spatial spectrum function, that is, the initial value of the direct wave estimated angle .

[0092] In this embodiment, in order to solve the problem that the mutual penetration between the transmit signal steering vector and the multipath steering vector of the transmit signal leads to the deterioration of the orthogonality between the noise subspace and the steering vector, and finally reduces the accuracy and reliability of the angle estimation, and even fails, the multipath effect is converted into a signal synthesized by the steering vectors, that is, equivalent to the direct wave signal and the reflected wave signal being added and synthesized, and in addition, due to the geometric relationship between the direct wave and the reflected wave angle when the positive angle is small, the expression of the reflected angle is as follows:

[0093] ;

[0094] wherein, represents the reflected angle, represents the height of the antenna array, represents the projection distance of the antenna array and the height measuring target on the horizontal plane;

[0095] Then, the steering vector of the received data can be re-expressed as follows:

[0096] ;

[0097] wherein, ​represents a steering vector of received data, represents a direct wave angle, represents a horizontal polarization wave reflection coefficient, represents a multipath steering vector of a transmitted signal, represents a reflection angle, represents a phase difference caused by a path difference between a direct wave and a reflected wave, represents a vector product, represents a steering vector of a return signal;

[0098] According to the redefined steering vector of received data, a covariance matrix is calculated and eigenvalue decomposition is performed to obtain a noise subspace, and in consideration of the horizontal polarization wave reflection coefficient Next, a MUSIC algorithm is used, and at this time, the MUSIC algorithm only needs to be related to Two-dimensional search, and the expression of the MUSIC is as follows:

[0099] ;

[0100] wherein, represents a search space spectrum value of a direct wave angle and a reflection coefficient; and an initial value of a horizontal polarization wave reflection estimation coefficient is obtained .

[0101] S3, according to the generalized multiple signal classification method and the multiple signal classification method, the estimated direct wave estimation angle initial value and the estimated horizontal polarization wave reflection estimation coefficient initial value estimated in S2 are alternately searched in order to obtain an estimated parameter group containing the estimated direct wave estimation angle and the estimated horizontal polarization wave reflection estimation coefficient.

[0102] In this embodiment, two parameters and are alternately searched in order of the initial value estimation, and an estimated parameter group K of the first estimation is obtained ; Since the calculation amount of the spectrum peak search of the MUSIC algorithm is huge, the alternately searching is used to reduce the workload of the search; The alternately searching is specifically as follows:

[0103] In the first search, a random direct wave angle is used as an initial value to estimate the horizontal polarization wave reflection coefficient, and an initial estimated value of the horizontal polarization wave transmission coefficient is obtained;

[0104] The initial estimated value of the horizontal polarization wave transmission coefficient is used as an initial value to estimate the direct wave angle, and an initial estimated value of the direct wave angle is obtained;

[0105] In the first search, a random direct wave angle is used as an initial value to estimate the horizontal polarization wave reflection coefficient, and an initial estimated value of the horizontal polarization wave transmission coefficient is obtained; K In the first search, a random direct wave angle is used as an initial value to estimate the horizontal polarization wave reflection coefficient, and an initial estimated value of the horizontal polarization wave transmission coefficient is obtained;​K Secondary estimate;

[0106] The first method using the emission coefficient of horizontally polarized waves K The first estimate is used to estimate the angle of the direct wave, thus obtaining the first estimate of the angle of the direct wave. K This involves multiple estimates to achieve alternating searches.

[0107] In this embodiment, when estimating the direct wave estimation angle, it is necessary to use generalized MUSIC to remove spurious peaks, as follows:

[0108] In the one-dimensional search using the MUSIC algorithm, redundant large spectral peaks and periodic small spectral peaks, i.e., false peaks, appeared. The phenomenon of lobe splitting in meter-wave MIMO radar was used to explain and resolve the issue of redundant spectral peaks, such as... Figure 3 As shown, it can be seen that the smaller the elevation angle of the target being measured, the more pronounced the influence of the periodic spectral peaks. Figure 4 As shown, 20lgP represents the logarithmic transformation of the search space spectrum value, and is the absolute value of the sum of the transmitted signal steering vector and the multipath steering vector when the horizontal polarization wave reflection coefficient is set to -0.9. ,in, The steering vector representing the transmitted signal The Middle m One element, Multipath steering vector representing the transmitted signal The Middle m One element; Figure 4 It can be seen that the peak amplitude decreases as the wavelength decreases and the antenna height increases. This indicates that radars with larger wavelengths (wider beams) and lower antenna heights are more prone to beam "grounding," leading to severe beam splitting. Therefore, in this embodiment, a smaller wavelength is selected while assuming that the antenna height is sufficiently high and less than the height of the target being measured. This effectively avoids the problem of getting trapped in local optima caused by redundant spectral peaks. However, sometimes the radar wavelength and antenna height are fixed during system design due to other factors and cannot be selected. Therefore, this invention uses a generalized MUSIC method that is not affected by the field emission coefficient to select the correct spectral peaks and avoid the occurrence of false redundant spectral peaks.

[0109] S4. Define the accuracy threshold and iteration threshold, and judge the estimated parameter group. If the accuracy of the estimated parameter group is less than the accuracy threshold or the number of iterations of the estimated parameter group is greater than the iteration threshold, output the direct wave estimated angle; otherwise, return to step S3.

[0110] In this embodiment, iterative precision is utilized. Define precision threshold The iteration threshold is defined using the maximum number of iterations; the estimated parameter set is judged using the accuracy threshold and the iteration threshold, and if the accuracy of the estimated parameter set is less than the accuracy threshold... Or the estimated number of iterations for the parameter set is greater than the iteration threshold. Output the estimated angle; otherwise, return to the alternating search step to obtain the estimated parameter value.

[0111] S5. The angle is estimated using the direct wave, and the target height is calculated by combining the projected distance between the antenna array and the target in the meter-wave multi-input multi-output radar multipath receiving signal model and the height of the antenna array, thus completing the multi-signal classification height measurement.

[0112] In this embodiment, the incident direct wave is taken as... Signal-to-noise ratio The experiment was conducted, and the mean squared error was defined as follows:

[0113] ;

[0114] in, Indicates the parameter to be measured. This represents the number of Monte Carlo random trials; 100 Monte Carlo trials are performed for each signal-to-noise ratio at the incident angle, with M=10 array elements and 10 snapshots. , , , as well as Under the conditions of MUSIC algorithm simulation, the following results were obtained: Figure 5 The mean square error of the angle measurement shown is as follows: Figure 6 The mean square error of altitude measurement shown shows that the angle measurement accuracy is higher under conditions of high signal-to-noise ratio and large elevation angle, while the angle measurement accuracy is lower in the low elevation angle region due to the characteristics of meter wave radar lobe splitting.

[0115] In this embodiment, under a relatively ideal model, the altitude measurement problem of meter-wave MIMO radar was analyzed, and the problem of redundant spectral peaks in the steering vector synthesis algorithm was discovered. A method for low-elevation-angle altitude measurement of meter-wave MIMO radar combining generalized MUSIC and alternating MUSIC searches was proposed. This method avoids the problem of a sharp increase in computational load caused by dense grid partitioning search, while achieving good low-elevation-angle altitude measurement results. The method of this invention does not require decoherence and has high engineering practical value. The signal model constructed by this invention is relatively ideal.

Claims

1. A multi-signal classification altimetry method based on meter-wave multi-input multi-output radar, characterized in that, Includes the following steps: S1. Based on the antenna array and the height measurement target, model the meter-wave multi-input multi-output radar multipath received signal, and perform matched filtering and vectorization on the received multi-input multi-output array received signal to calculate the output signal; S2. Calculate the array data covariance based on the output signal and combine them into a covariance matrix. Obtain the noise subspace based on the covariance matrix. Use the generalized multiple signal classification method to obtain the initial value of the direct wave and use multiple signal classification to obtain the initial value of the reflection coefficient of the horizontal polarized wave. S3. Based on the generalized multiple signal classification method and the multiple signal classification method, perform alternating searches in the order of estimating the initial value of the direct wave estimated angle and estimating the initial value of the horizontal polarized wave reflection coefficient in S2 to obtain the estimated parameter set containing the direct wave estimated angle and the horizontal polarized wave reflection coefficient. The steps of the alternating search are as follows: In the first search, the reflection coefficient of the horizontally polarized wave is estimated using the random direct wave angle as the initial value, and the initial estimated value of the emission coefficient of the horizontally polarized wave is obtained. The initial estimate of the horizontally polarized wave emission coefficient is used as the initial value to estimate the angle of the direct wave, thus obtaining the initial estimate of the angle of the direct wave. Except for the first search, the reflection coefficient of the horizontally polarized wave is estimated using the current estimate of the direct wave angle, and the next estimate of the emission coefficient of the horizontally polarized wave is obtained. The angle of the direct wave is estimated by using the next estimate of the horizontal polarization wave emission coefficient, thus obtaining the next estimate of the angle of the direct wave and realizing alternating search. S4. Define the accuracy threshold and the iteration threshold, and judge the estimated parameter group. If the accuracy of the estimated parameter group is less than the accuracy threshold or the number of iterations of the estimated parameter group is greater than the iteration threshold, output the direct wave estimated angle; otherwise, return to step S3. S5. The angle is estimated using the direct wave, and the target height is calculated by combining the projected distance between the antenna array and the target in the meter-wave multi-input multi-output radar multipath receiving signal model and the height of the antenna array. This completes the multi-signal classification height measurement.

2. The multi-signal classification altimetry method based on meter-wave multi-input multi-output radar according to claim 1, characterized in that, S1 includes the following steps: S101. Based on the antenna array and the altitude measurement target, model the multipath received signal of the meter-wave multi-input multi-output radar. S102. Based on the transmitted signal steering vector and the transmitted signal multipath steering vector in the meter-wave multi-input multi-output radar multipath receiving signal model, the signal of the height measurement target being illuminated by the radar is obtained. S103. Based on the radar illumination signal of the target and the array elements of the antenna array, obtain the echo signal of the corresponding array element, and use the echo signal of each array element to obtain the multi-input multi-output array received signal. S104. Perform matched filtering and vectorization on the received signal of the multi-input multi-output array to calculate the output signal.

3. The multi-signal classification altimetry method based on meter-wave multi-input multi-output radar according to claim 1, characterized in that, S2 includes the following steps: S201. Calculate the array data covariance based on the output signal, combine the array data covariance into a covariance matrix, and perform eigenvalue decomposition on the covariance matrix to obtain the noise subspace. S202. Based on the noise subspace, the initial value of the direct wave estimation angle is obtained using the generalized multiple signal classification method. S203. Based on the meter-wave multi-input multi-output radar multipath receiving signal model, the steering vector of the received data is calculated using the horizontal polarization wave reflection coefficient and the direct wave angle. Based on multiple signal classification, the initial value of the horizontal polarization wave reflection estimation coefficient is obtained using the steering vector of the received data and the noise subspace.

4. The multi-signal classification altimetry method based on meter-wave multi-input multi-output radar according to claim 3, characterized in that, The expression for the multiple signal classification is as follows: in, The search space spectrum values ​​represent the angle of arrival and the reflection coefficient. Indicates the guide vector for receiving data. Indicates the angle of the direct wave. Represents the reflection coefficient of horizontally polarized waves. Represents the noise subspace. Indicates the steering vector of the transmitted signal. This represents the multipath steering vector of the transmitted signal. Indicates the reflection angle. This represents the phase difference caused by the path difference between the direct wave and the reflected wave. Represents the vector product. Represents the echo signal steering vector. Indicates the height of the antenna array. This indicates the projected distance between the antenna array and the target being measured on the horizontal plane.

5. The multi-signal classification altimetry method based on meter-wave multi-input multi-output radar according to claim 1, characterized in that, The specific steps of S4 are as follows: Define the precision threshold using the iteration precision, and define the iteration threshold using the maximum number of iterations; The estimated parameter set is judged using the accuracy threshold and the iteration threshold. If the accuracy of the estimated parameter set is less than the accuracy threshold or the number of iterations of the estimated parameter set is greater than the iteration threshold, the direct wave estimated angle is output; otherwise, return to step S3.

6. The multi-signal classification altimetry method based on meter-wave multi-input multi-output radar according to claim 1, characterized in that, The expression for calculating the target height is as follows: in, Indicates the target height. This indicates the projected distance between the antenna array and the altimeter target. Indicates the estimated angle of the direct wave. This indicates the height of the antenna array.

Images