A Subband Synthesis Method Based on Scattering Center Classification

Abstract

The present invention provides a sub-band synthesis method based on scattering center classification. First, the linear phase between sub-bands is estimated and compensated. Then, scattering centers are extracted from the sub-band echoes after compensating the linear phase, and the scattering centers in each sub-band are divided into inherent scattering centers and unique scattering centers according to the relative distance parameters of the scattering centers. For the inherent scattering centers, a new fixed phase estimation and compensation method is introduced to compensate for the additional phase difference and the original fixed phase difference introduced by the fluctuations of the scattering characteristics. Then, corresponding full-band echoes are generated for different types of scattering centers and superimposed, and finally the target full-band echo is obtained. The present invention fully considers the characteristic that the electromagnetic scattering characteristics of the target have fluctuations in the ultra-wideband frequency band range, and introduces a new fixed phase estimation and compensation method to compensate for the additional phase term brought by the fluctuations of the target scattering characteristics, thereby improving the final accuracy of sub-band synthesis.

Description

Technical Field

[0001] The present invention belongs to the technical field of radar signal processing, and particularly relates to a sub-band synthesis method based on the GTD model and the Root-MUSIC algorithm. Background Art

[0002] Based on radar theory, the range resolution of a target depends on the bandwidth of the radar signal. The relative bandwidth of the traditional radar transmitted signal is usually less than 10% of the carrier frequency, so it is difficult to achieve a high ranging resolution. To address the limitation of the radar bandwidth, there are two methods to obtain an ultra-wideband (UWB) radar signal. One method is to directly construct UWB radar hardware, which will lead to a significant increase in system complexity and design cost. Another effective method for generating UWB signals is the sub-band synthesis technique. It is a signal processing method that fuses the echo signals of two or more radar sub-bands into an ultra-wideband radar echo signal, thus avoiding the high cost of building a true UWB radar and further improving the range resolution of radar targets.

[0003] With the development of sub-band synthesis technology, there are linear and fixed phases between sub-bands, which need to be estimated and compensated. Sub-band synthesis methods can be classified into two categories: model-based algorithms, including traditional spectral estimation algorithms and sparse representation algorithms, and data-driven algorithms. Spectral estimation algorithms are widely used in sub-band synthesis processing due to their low computational complexity. By analyzing the existing sub-band synthesis methods, it can be seen that the existing methods all assume that the electromagnetic scattering characteristics of the target remain consistent throughout the synthesis frequency band, thus obeying the electromagnetic scattering model under the same parameters. However, in actual situations, the electromagnetic scattering characteristics of the target fluctuate within the ultra-wideband frequency range, which will affect the final accuracy of sub-band synthesis. Summary of the Invention

[0004] The technical problem to be solved by the present invention is, aiming at the problem that the scattering characteristics of the target fluctuate within the ultra-wideband frequency domain, thus affecting the sub-band synthesis accuracy, by analyzing the causes of the fluctuations, and then a sub-band synthesis method for improving the accuracy of sub-band synthesis is proposed.

[0005] To solve the above technical problem, based on the GTD model, the phase mismatch situation between multiple sub-bands is carefully modeled, and it is deduced that the fluctuations of the electromagnetic scattering characteristics of the target will introduce additional phase differences on the basis of the original fixed phases between sub-bands, thus reducing the accuracy of sub-band synthesis. The technical solution adopted for this analysis is a sub-band synthesis method based on scattering center classification to estimate and compensate the additional phase differences introduced by the fluctuations of the target scattering characteristics, thereby improving the accuracy of sub-band synthesis. The specific steps are as follows:

[0006] Estimate and compensate the linear phase between sub-bands;

[0007] Extract the electromagnetic scattering center parameters of the target based on the sub-band echo after compensating the linear phase;

[0008] Divide the scattering centers in each sub-band into inherent scattering centers and unique scattering centers, and respectively enter the inherent scattering center processing step or the unique scattering center processing step;

[0009] Inherent scattering center processing step: Calculate a new fixed phase based on the inherent scattering center and perform compensation to compensate for the additional phase difference introduced by the scattering characteristic fluctuations and the original fixed phase difference, and restore the phase correlation of the inherent scattering centers between sub-bands; Then, based on each inherent scattering center after restoring the phase correlation, perform sub-band echo reconstruction to obtain the reconstructed sub-band echo of the inherent scattering center; Finally, perform sub-band synthesis on the reconstructed sub-band echo of the inherent scattering center to obtain the full-band echo of the inherent scattering center;

[0010] Unique scattering center processing step: Perform sub-band extrapolation on the unique scattering centers in each sub-band respectively to generate the full-band echo of the unique scattering centers;

[0011] Sub-band synthesis step: Superimpose the full-band echo of the inherent scattering center and the full-band echo of the unique scattering center to obtain the superimposed full-band echo, and then use the original sampling data of each sub-band and the blank frequency band between sub-bands to form the final full-band echo of the target using the corresponding parts of the superimposed full-band echo.

[0012] Specifically, extract the electromagnetic scattering center parameters of the target based on the GTD model and the Root-MUSIC algorithm. The electromagnetic scattering center parameters include the order of the GTD model, the scattering intensity of the scattering center, and the relative distance.

[0013] The present invention first considers the characteristic that the linear phase existing between sub-bands will cause the one-dimensional range profile of the sub-band to shift left and right, and first estimates and compensates the linear phase between sub-bands. Once the linear phase between sub-bands is compensated, the peak points of the one-dimensional range profile of the sub-band will be aligned and correspond one by one. Then, based on the GTD model, extract the scattering centers from the sub-band echo after compensating the linear phase, and divide the scattering centers in each sub-band into two categories according to the relative distance parameter of the scattering center: the inherent scattering center with the relative distance parameter within plus or minus one distance resolution and the unique scattering center with the relative distance parameter not within plus or minus one distance resolution. A new fixed phase estimation and compensation method is introduced for the inherent scattering center to compensate for the additional phase difference introduced by the scattering characteristic fluctuations and the original fixed phase difference. Finally, perform synthesis on different categories of scattering centers respectively to generate the corresponding full-band echo, and superimpose them to obtain the final ultra-wideband echo.

[0014] The beneficial effects of the present invention are as follows: fully considering the characteristic that the target electromagnetic scattering characteristics fluctuate within the ultra-wideband frequency band, a new fixed phase estimation and compensation method is introduced to compensate for the additional phase term caused by the fluctuation of the target scattering characteristics, thereby improving the final accuracy of sub-band synthesis. Description of the Drawings

[0015] Figure 1 It is a block diagram of the sub-band synthesis signal processing flow.

[0016] Figure 2 It is a block diagram of the improved sub-band synthesis method based on the GTD model and the Root-MUSIC algorithm in the embodiment.

[0017] Figure 3 It is the experimental target.

[0018] Figure 4 It is a comparison of the accuracy between the present invention and the traditional sub-band synthesis method. Detailed Embodiment

[0019] The following clearly and completely describes the detailed embodiment of the present invention with reference to the drawings, so that those skilled in the art can better understand the invention.

[0020] The Geometrical Theory of Diffraction (GTD) model is a high-frequency approximation theory in the analysis of electromagnetic wave propagation and scattering, and is an important tool in computational electromagnetics.

[0021] Figure 1 The following shows three key steps of the traditional sub-band synthesis method: The first step is sub-band coherent processing, that is, estimating and compensating the linear phase difference and fixed phase difference between the echoes of each sub-band. After compensation, the phases between the sub-bands are aligned. The second step is the estimation of the GTD model parameters in the full frequency band. The third step is to estimate the full-frequency echo signal of the target according to the GTD model parameters in the full frequency band and the original sub-band echo data. Interpolation is performed between sub-band 1 and sub-band 2, and extrapolation between sub-band 1 and sub-band 2 to obtain the full-frequency echo. In the first step of the traditional sub-band synthesis method, it is generally assumed that the scattering characteristics of different sub-bands are approximately the same, that is, different sub-bands conform to a unified GTD parameter model. Therefore, the phase difference between sub-bands only has a linear phase and a fixed phase.

[0022] However, when the electromagnetic scattering characteristics of the target fluctuate within the ultra-wideband frequency range, the sub-bands will no longer conform to a unified GTD parameter model.

[0023] Based on the basic radar theory and the GTD model, the sub-band echo can be expressed as:

[0024] ;

[0025] In the above formula, n is the sampling point serial number, and the sub-band serial number i = 1, 2, represents the representation of the echo of sub-band i at the nth sampling point, represents the representation of the echo of sub-band 2 at the nth sampling point, is the GTD model order of sub-band i, represents the amplitude of the echo of sub-band i at the mth scattering center, represents the phase of the echo of sub-band i at the mth scattering center, represents the frequency-dependent factor of the echo of sub-band i at the mth scattering center, is the relative distance of the echo of sub-band i at the mth scattering center. The relative distance is the distance between the scattering center and the radar minus the distance between the reference point and the radar. Generally, the actual distance of the radar-detected target is taken. For example, if the target is at 1000 m and there are three scattering centers on the target at 905 m, 1001 m, and 1005 m respectively, then the relative distances are -5 m, 1 m, and 5 m; is the speed of light, is the starting frequency, is the frequency sampling interval, is the pole of the mth scattering center of the sorted echo of sub-band i, is the amplitude of the mth scattering center of the sorted echo of sub-band i. exp is the natural exponential function, and j is the imaginary unit. The amplitude of the scattering center is a complex amplitude. At this time, taking sub-band 1 as a reference, there is a linear phase and a fixed phase between sub-band 1 and sub-band 2, as shown in the following formula:

[0026] ;

[0027] and represent the number of sampling points of sub-band 1 and sub-band 2 respectively, is the number of sampling points of the full-band echo. The poles and amplitudes of the scattering centers of sub-band 1 and sub-band 2 are as follows respectively:

[0028] ;

[0029] When the scattering characteristics of the target fluctuate, the amplitude, phase, and frequency-dependent factor of the scattering center under the same relative distance parameters will no longer remain unchanged. Comparing the phase term of the scattering center amplitude , it can be seen that if and are changing, an additional phase difference will be introduced and superimposed on the original fixed phase . The additional phase term is as shown in the following formula:

[0030] ;

[0031] That is, the new fixed phase is:

[0032] ;

[0033] It can be deduced from the above that the change in the scattering characteristics of the target will introduce an additional phase term on the original fixed phase to form a new fixed phase, and each scattering point is different.

[0034] The Root-MUSIC algorithm is an improved version of the classical MUSIC (Multiple Signal Classification) algorithm. By replacing spectral search with polynomial root finding, it significantly improves the computational efficiency and parameter estimation accuracy.

[0035] Now, based on the Root-MUSIC algorithm, the linear phase and the new fixed phase of the sub-band echo considering the fluctuations of the target electromagnetic scattering characteristics are estimated and compensated. The detailed steps are as follows:

[0036] Step 1: Estimation and compensation of the linear phase based on the principle of minimum entropy: Assume that the true linear phase between sub-bands is , represents the linear phase to be estimated and compensated. is the echo of sub-band 2 after compensating . The inverse Fourier transform of the echo of sub-band 2 after compensating obtains the high-resolution one-dimensional range image of sub-band 2. The inverse Fourier transform of the echo of sub-band 1 obtains the high-resolution one-dimensional range image of sub-band 1 as . Their amplitudes are superimposed to obtain the one-dimensional range image amplitude sequence after linear phase compensation between sub-bands:

[0037] ;

[0038] Its amplitude distribution is:

[0039] ;

[0040] Calculate the information entropy :

[0041] ;

[0042] When the compensated linear phase is closer to the true linear phase , the one-dimensional range image coincidence degree of the two sub-bands is the highest, corresponding to the minimum information entropy. Therefore, the information entropy to be compensated can be obtained by minimizing the following formula:

[0043] ;

[0044] At this time, the obtained by compensation estimation is fed back to sub-band 2 to obtain the compensated echo of sub-band 2 after that .

[0045] Step 2: Scattering center parameter extraction and classification: After estimating and compensating the linear phase, based on the GTD model and the Root-MUSIC algorithm, the scattering centers of the target are extracted from the sub-band echoes and . Then, based on the relative distance parameter of the scattering centers, the scattering centers are classified. The classification criterion is: the scattering centers with relative distance parameters within plus or minus one range resolution cell are the sub-band intrinsic scattering centers ISC, and those exceeding plus or minus one range resolution are the unique scattering centers USC. At this time, the poles and amplitudes of the scattering centers: are also correspondingly divided into two categories: and .

[0046] In the traditional sub-band synthesis method, the fixed phase estimation is calculated as shown in the following formula:

[0047] ;

[0048] where angle is the phase angle function; , there are two problems in the above calculation. One is the one-to-one correspondence problem of the scattering centers, and the other is that the influence of the fluctuation of the target scattering characteristics is not considered. Because the fluctuation of the scattering characteristics will introduce an additional phase term: . Therefore, for each scattering center, the fixed phase is different. For this reason, fixed phase compensation needs to be performed for each scattering center.

[0049] Step 3: For the intrinsic scattering centers, calculate the fixed phase difference as shown in the following formula: The calculation method at this time includes both the original fixed phase and the additional phase term introduced by the fluctuation of the target scattering characteristics :

[0050] ;

[0051] For the intrinsic scattering centers, compensate the estimated new fixed phase:

[0052] ;

[0053] Step 4: Based on the poles of the inherent scattering centers after restoring the phase correlation and the amplitudes of the scattering centers , reconstruct the sub-band echoes of the inherent scattering centers and :

[0054] ;

[0055] where is the number of inherent scattering centers.

[0056] Step 5: Perform joint parameter estimation on the echoes of the inherent scattering centers in the above formula to obtain the poles of the scattering centers in the full frequency band and the amplitudes so as to obtain the full frequency band echoes of the inherent scattering centers

[0057] ;

[0058] Step 6: For the unique scattering centers, perform 2-fold frequency extrapolation to obtain the full frequency band echoes of the unique scattering centers in sub-band 1 and sub-band 2 and :

[0059] ;

[0060] Step 7: Superimpose the full frequency band echoes of the inherent scattering centers and the full frequency band echoes of the unique scattering centers in each sub-band to obtain the superimposed full frequency band echoes :

[0061] ;

[0062] To reduce errors, using the original sampling data of sub-band 1 and sub-band 2, use the corresponding part of the full frequency band echo in the blank frequency band between sub-band 1 and sub-band 2, and finally obtain the target full frequency band echo as follows:

[0063] ;

[0064] Thus, the improved sub-band synthesis method based on the GTD model and the Root-MUSIC algorithm is completed. Next, combined with the actual sub-band parameters, each sub-band is synthesized according to the above method to compare the accuracy of sub-band synthesis, so as to reflect the improvement of the sub-band synthesis accuracy by the present invention.

[0065] First, set the sub-band synthesis full frequency range as: 4.5 - 8.5 GHz, sub-band 1 as 4.5 - 5.5 GHz, sub-band 2 as: 7.5 - 8.5 GHz, and the blank frequency band as: 5.5 - 7.5 GHz. Both radars use linear frequency modulation signals as radar transmission signals, with a bandwidth of 1 GHz and a pulse of 10 us. The detection target is as Figure 3 shown. Its scattered field data is measured in a microwave anechoic chamber. Based on the scattered field data of the target and the radar transmission signal, the sub-band echoes and the ideal full-frequency echo of the target are generated. At the same time, set the original linear phase and fixed phase between sub-bands as: , .

[0066] Based on Figure 2 the steps 1 - 7 shown, signal processing is performed on the sub-band echoes to obtain the full-frequency echo after sub-band synthesis. And a correlation coefficient is introduced to measure the accuracy of sub-band synthesis, as shown below, is the ideal full-frequency echo, is the full-frequency echo obtained by sub-band synthesis. At the same time, it can be seen from the following formula that the larger the correlation coefficient, the higher the accuracy of sub-band synthesis, and its maximum value is 1.

[0067] ;

[0068] is the 2-norm. Compared with the traditional method, the comparison results of the correlation coefficients are as Figure 4 shown. The correlation coefficients of the method proposed in the present invention are all higher than 0.9, which has been greatly improved compared with the traditional method. For example, at a signal-to-noise ratio of 15 dB, the correlation coefficient of the present invention is 25% higher than that of the traditional method. It can be seen from the above that the present invention fully considers the influence of the fluctuations of the target scattering characteristics on sub-band synthesis, and the new fixed phase difference estimation and compensation method based on the Root-MUSIC algorithm effectively improves the accuracy of sub-band synthesis.

[0069] Although the above describes the illustrative specific embodiments of the present invention for the convenience of those skilled in the art to understand the present invention, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention defined and determined by the appended claims, these changes are obvious, and all inventions created using the concept of the present invention are within the scope of protection.

Claims

1. A sub-band synthesis method based on scattering center classification, characterized in that, It includes the following steps: Estimate and compensate the linear phase between sub-bands; Extract the electromagnetic scattering center parameters of the target based on the sub-band echo after compensating the linear phase; Divide the scattering centers in each sub-band into inherent scattering centers and unique scattering centers, and respectively enter the inherent scattering center processing step or the unique scattering center processing step; Inherent scattering center processing step: Calculate a new fixed phase based on the inherent scattering center and compensate it to compensate for the additional phase difference and the original fixed phase difference introduced by the scattering characteristic fluctuations, and restore the phase correlation of the inherent scattering centers between sub-bands; Then, based on each inherent scattering center after restoring the phase correlation, perform sub-band echo reconstruction to obtain the reconstructed sub-band echo of the inherent scattering center; Finally, perform sub-band synthesis on the reconstructed sub-band echo of the inherent scattering center to obtain the full-band echo of the inherent scattering center; Unique scattering center processing step: Perform sub-band extrapolation on the unique scattering centers in each sub-band respectively to generate the full-band echo of the unique scattering center; Sub-band synthesis step: Superimpose the full-band echo of the inherent scattering center and the full-band echo of the unique scattering center to obtain the superimposed full-band echo, and then use the original sampling data of each sub-band and the blank frequency band between sub-bands to form the final full-band echo of the target with the corresponding part of the superimposed full-band echo; Among them, the scattering centers with relative distance parameters within plus or minus one range resolution cell are inherent scattering centers, and those exceeding plus or minus one range resolution are unique scattering centers, and the relative distance is the distance between the scattering center and the radar minus the distance between the reference point and the radar.

2. The method according to claim 1, wherein Estimate and compensate the linear phase between sub-bands based on the minimum entropy theorem.

3. The method according to claim 2, wherein Estimating the linear phase between subbands based on the minimum entropy theorem The specific method is as follows: ； where argmin is the minimum solution function, represents the linear phase variable, , ; is the amplitude distribution of the one-dimensional range profile amplitude sequence after linear phase compensation between sub-bands; is the information entropy of, and N is the number of sampling points of the full-band echo.

4. The method according to claim 1, characterized in that, Extract the electromagnetic scattering center parameters of the target based on the geometric theory of diffraction GTD model and the Root-MUSIC algorithm. The electromagnetic scattering center parameters include the order of the GTD model, the scattering amplitude and relative distance of the scattering center.

5. The method according to claim 4, wherein Divide the scattering centers in the sub-band into inherent scattering centers and unique scattering centers based on the relative distance of the electromagnetic scattering centers in each sub-band.

6. The method according to claim 5, wherein The scattering centers with relative distance within plus or minus one range resolution are inherent scattering centers; The scattering centers with relative distance exceeding plus or minus one range resolution are unique scattering centers.

7. The method according to claim 4, characterized in that The method for calculating a new fixed phase based on the inherent scattering center is: Calculate the additional phase term introduced by the fluctuation of the target scattering characteristics at the m-th inherent scattering center : ； where angle is the phase angle function, is the amplitude of the echo of sub-band 2 at the m-th intrinsic scattering center, is the amplitude of the echo of sub-band 1 at the m-th intrinsic scattering center; At the m-th intrinsic scattering center, the amplitude after obtaining the new fixed phase after compensation is as follows: ； Where exp is the natural exponential function and j is the imaginary unit.

8. The method according to claim 7, wherein Reconstruct the sub - band echo of the intrinsic scattering center based on the poles and the amplitude of the m - th intrinsic scattering center of sub - band 1 and sub - band 2 after restoring the phase correlation and : ； Among them, is the number of inherent scattering centers, and respectively represent the number of sampling points in sub-band 1 and sub-band 2, and N is the number of sampling points of the full-band echo, is the amplitude of the sub-band 1 echo at the m-th inherent scattering center, is the amplitude of the sub-band 2 echo after compensation at the m-th inherent scattering center, and are the poles of the sub-band 1 echo and the sub-band 2 echo at the m-th inherent scattering center, and n is the sampling point sequence number.

Images