A method for azimuth wide-beam distributed SAR spectrum reconstruction

Abstract

This invention belongs to the field of synthetic aperture radar signal processing and spectrum reconstruction technology, and particularly relates to an azimuth wide-beam distributed SAR spectrum reconstruction method. After quantitatively analyzing the imaging errors of traditional spectrum reconstruction algorithms applied to azimuth wide-beam distributed SAR systems, an improved azimuth wide-beam distributed SAR spectrum reconstruction algorithm is proposed. This reconstruction algorithm improves the spectrum reconstruction accuracy by constructing a reconstruction function that includes azimuth spatial variation channel mismatch during spectrum reconstruction. Furthermore, by performing matrix decomposition on the reconstruction function, range spatial variation and azimuth spatial variation are decoupled, reducing the computational load of spectrum reconstruction and achieving fast and high-precision spectrum reconstruction of azimuth wide-beam distributed SAR.

Description

Technical Field

[0001] This invention belongs to the field of synthetic aperture radar signal processing and spectrum reconstruction technology, and particularly relates to an azimuth wide-beam distributed SAR spectrum reconstruction method. Background Technology

[0002] Synthetic Aperture Radar (SAR) is a high-resolution radar system capable of scene imaging. Due to its advantages such as all-weather, all-time operation, long range, and wide mapping swath, SAR has received widespread attention since its inception in the 1950s and has been widely applied in both military and civilian fields. Multi-channel SAR systems are SAR systems with multiple transmit or receive channels. Compared to single-channel SAR, they possess higher degrees of freedom, effectively improving system performance and expanding the application range of SAR systems. They play an irreplaceable role in high-resolution wide-swath imaging, SAR ground moving target indication, and topographic mapping. Traditional SAR systems form multi-channel SAR systems by dividing the entire antenna array into multiple subarrays, i.e., centralized multi-channel SAR. With the increasing demands of SAR applications, centralized multi-channel SAR systems are becoming increasingly complex. To avoid the problems of centralized multi-channel SAR systems, researchers have proposed distributed SAR.

[0003] High-resolution wide-swath imaging, as an important development direction of distributed SAR, is a hot research area. Resolution and swath width are key performance indicators of SAR systems. For single-channel SAR systems, regardless of the operating mode, the minimum antenna area limitation creates a contradiction in system design, making it difficult to simultaneously achieve high azimuth resolution and wide-swath mapping. Distributed SAR can overcome the minimum antenna area limitation and simultaneously obtain high azimuth resolution and wide mapping width. Key technologies for high-resolution wide-swath imaging in distributed SAR include: signal error estimation and compensation, spectrum reconstruction, and SAR imaging. SAR imaging processing is basically the same as that of traditional single-channel SAR. However, compared to traditional centralized azimuth multi-channel SAR, distributed SAR faces challenges in high-resolution wide-swath imaging, such as channel error estimation caused by cross-trajectory baselines and difficulties in spectrum reconstruction due to non-Nyquist sampling. Summary of the Invention

[0004] The technical problem solved by this invention is to overcome the shortcomings of the prior art and propose an azimuth wide-beam distributed SAR spectrum reconstruction method. This method can solve the problems of non-negligible azimuth spatial variation in channel mismatch in azimuth wide-beam distributed SAR, blurred imaging by traditional transfer function method, and the inability to compensate for channel mismatch by simply multiplying each channel by the conjugate of the channel mismatch, which requires processing multiple channels together.

[0005] The technical solution of this invention is:

[0006] A method for azimuth wide-beam distributed SAR spectrum reconstruction, the steps of which include:

[0007] S1. The original echo data is compressed by range and registered by range direction through matched filtering. The purpose of range compression is to localize the channel mismatch. The purpose of range registration is to estimate and compensate for the inter-channel envelope movement caused by the cross-track baseline and the distance sampling time delay based on the range-frequency cross-correlation method, so as to obtain the range-registered data.

[0008] S2, perform range-space-variable phase mismatch compensation on the range-registered data obtained in step S1 based on the inertial navigation data to obtain the mismatch-compensated data. The specific method is as follows: after calculating the baseline between channels through the inertial navigation data, the corresponding range-space-variable phase mismatch is obtained, and range-space-variable phase mismatch compensation is performed in each channel by conjugate multiplication to obtain the mismatch-compensated data.

[0009] S3, perform deviation correction on the phase error estimation and compensation of the OSM channel of the mismatch-compensated data obtained in step S2 to obtain the error-compensated data;

[0010] S4. Process the error-compensated data obtained in step S3. The specific method is as follows: First, calculate the imaging error when azimuth spatial variation is ignored, and judge the imaging error. When the imaging error is less than or equal to the set threshold, the imaging error can be ignored. Then, the transfer function method is used to reconstruct the spectrum of the error-compensated data. When the imaging error is greater than the set threshold, the imaging error cannot be ignored. Then, the corresponding azimuth channel mismatch needs to be compensated.

[0011] S5. Perform SAR imaging on the unambiguous signal Doppler spectrum obtained in step S4 after azimuth channel mismatch or spectrum reconstruction. Specifically, use the range Doppler algorithm or linear frequency modulation algorithm for imaging processing to finally obtain an unambiguous SAR image.

[0012] In step S4, when the imaging error is not negligible, it is necessary to compensate for the azimuth channel mismatch. The specific method is as follows: First, the transfer function of the linear system composed of distributed SAR is expressed as:

[0013] H(r,f a )=B(f a )P(r,f a (1)

[0014] in:

[0015]

[0016] P(r,f a )=[p1(r,fa ),p2(r,f a ),...,p N (r,f a (3)

[0017]

[0018] M represents the number of channels, and N represents the Doppler ambiguity number. To achieve error estimation, M > N. `diag{}` represents a diagonal matrix with the elements within the brackets arranged diagonally. The phase mismatch is caused by the phase error of channel m and the channel position measurement error, where m = 1, 2, ..., M, y m The y-axis coordinate of the equivalent phase center of the m-th channel receiving unit, v p Let λ be the velocity of the flight platform, λ be the carrier wavelength, r be the target slant range, and f be the velocity of the flight platform. a f is the Doppler frequency. prf Indicates the pulse repetition frequency, ΔR″ m,i (r,f a The residual slant range error between channel m and the reference channel due to the cross-trajectory baseline and azimuth beam variation in two dimensions can be obtained using the above formula. -1 (f a ).

[0019] To further improve processing efficiency, ΔR″ can be calculated. m,i (r,f a The imaging error caused by spatial variation of distance, if the ambiguity component produced by this approximation is less than the system's azimuth ambiguity ratio (AASR) requirement within a certain range, is considered to be within this range. m,i (r,f a The distance variation is negligible and can be determined by ΔR″. m,i (r0,f a Let r0 represent the slant distance from the center of the target scene. At this point, the signal steering matrix P(r,f) a ) can be approximated as P′(f) a The transfer function of a linear system composed of distributed SAR can be further simplified to:

[0020] H′(f a )=B(f a )P′(f a (5)

[0021] Wherein, P′(f a ) is to add ΔR″ in equation (4) m,i (r,f a Replace ) with ΔR″ m,i (r0,f a P(r,f)a Let H' be a column vector. Using the above equation, the simplified reconstruction filter can be obtained as H''. -1 (f a ), using H′ -1 (f a This allows for distributed SAR spectrum recovery under the azimuth wide-beam assumption.

[0022] Beneficial effects

[0023] (1) In the method of the present invention, the distance spatial phase mismatch compensation based on inertial navigation data provides more independent and identically distributed units for subsequent processing.

[0024] (2) In the method of the present invention, the azimuth wide-beam distributed SAR spectrum reconstruction algorithm constructs a reconstruction function that includes azimuth spatial variation channel mismatch, and compensates for azimuth spatial variation channel mismatch during spectrum reconstruction, thereby improving the spectrum reconstruction accuracy.

[0025] (3) In the method of the present invention, in ΔR″ m,i (r,f a When the imaging error caused by spatial variation of distance is within a set threshold range and the blur component generated by this approximation is less than the azimuth blur ratio (AASR) requirement of the system, ΔR″ can be ignored. m,i (r,f a The distance spatial variation significantly reduces the computational load without sacrificing reconstruction performance.

[0026] (4) In the method of the present invention, ΔR″ m,i (r,f a When the range spatial variation is negligible, the reconstruction function is matrix decomposed, which decouples the range spatial variation and azimuth spatial variation, reduces the computational load of spectrum reconstruction, realizes fast and high-precision spectrum reconstruction of azimuth wide-beam distributed SAR, and improves the azimuth ambiguity suppression capability of the transfer function method for azimuth wide-beam distributed SAR systems.

[0027] (5) In the method of the present invention, the reconstructed azimuth wide-beam distributed SAR spectrum is imaged to obtain a more accurate SAR image. Attached Figure Description

[0028] Figure 1 A flowchart of azimuth wide-beam distributed SAR high-resolution wide-swath imaging process;

[0029] Figure 2 Optical map of the imaging scene;

[0030] Figure 3(a) shows the imaging results after spectral reconstruction using the traditional method;

[0031] Figure 3(b) shows the imaging results after spectral reconstruction using the method of the present invention;

[0032] Figure 4(a) shows the azimuth section of the traditional method imaging result for target A;

[0033] Figure 4(b) shows the azimuth section of the imaging result of target B using the traditional method;

[0034] Figure 4(c) Azimuth section of the imaging result of target A using the proposed method;

[0035] Figure 4(d) shows the azimuth section of the imaging result of target B using the proposed method. Detailed Implementation

[0036] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0037] Azimuth wide-beam distributed SAR high-resolution wide-swath imaging includes the following steps:

[0038] S1: Range compression and range orientation registration;

[0039] S2: Range spatially variable phase mismatch compensation based on inertial navigation data;

[0040] S3: Error estimation of orthogonal subspace projection channels based on deviation correction;

[0041] S4: Phase Error Estimation and Compensation for Range Spatial Variation

[0042] S5: Unambiguous spectral reconstruction and SAR imaging processing based on an improved azimuth spectrum reconstruction function.

[0043] Example

[0044] like Figure 1 As shown, azimuth wide-beam distributed SAR high-resolution wide-swath imaging includes the following steps:

[0045] S1. The original echo data is compressed by range and registered by range direction through matched filtering. The purpose of range compression is to localize the channel mismatch. The purpose of range registration is to estimate and compensate for the inter-channel envelope movement caused by the cross-track baseline and the distance sampling time delay based on the range-frequency cross-correlation method, so as to obtain the range-registered data.

[0046] S2, Based on the inertial navigation data, perform range-space-variable phase mismatch compensation on the range-registered data obtained in step S1 to obtain mismatch-compensated data. The specific method is as follows: After calculating the baseline between channels through inertial navigation data, the corresponding range-space-variable phase mismatch is obtained, and range-space-variable phase mismatch compensation is performed in each channel by conjugate multiplication to obtain mismatch-compensated data.

[0047] S3, perform deviation correction on the phase error estimation and compensation of the OSM channel of the compensated data obtained in step S2 to obtain the error-compensated data;

[0048] S4. Perform spectrum reconstruction on the error-compensated data obtained in step S3. The specific method is as follows: First, calculate the imaging error when azimuth spatial variation is ignored, and judge the imaging error. When the imaging error is less than or equal to the set threshold, the imaging error can be ignored. Then, the transfer function method is used to perform spectrum reconstruction on the error-compensated data. When the imaging error is greater than the set threshold, the imaging error cannot be ignored. Then, the corresponding azimuth channel mismatch needs to be compensated.

[0049] S5. Perform SAR imaging on the unambiguous signal Doppler spectrum obtained from the azimuth channel mismatch and spectrum reconstruction in step S4. Specifically, use the traditional range Doppler algorithm or linear frequency modulation algorithm for imaging processing to finally obtain an unambiguous SAR image.

[0050] In step S4, when the imaging error is not negligible, it is necessary to compensate for the azimuth channel mismatch. The specific derivation formula and process are as follows. First, the transfer function of the linear system composed of distributed SAR can be expressed as:

[0051] H(r,f a )=B(f a )P(r,f a (1)

[0052] in:

[0053]

[0054] P(r,f a )=[p1(r,f a ),p2(r,f a ),...,p N (r,f a (3)

[0055]

[0056] M represents the number of channels, and N represents the Doppler ambiguity number. To achieve error estimation, M > N. `diag{}` represents a diagonal matrix with the elements within the brackets arranged diagonally. The phase mismatch is caused by the phase error of channel m and the channel position measurement error, where m = 1, 2, ..., M, y m The y-axis coordinate of the equivalent phase center of the m-th channel receiving unit, v p Let λ be the velocity of the flight platform, λ be the carrier wavelength, r be the target slant range, and f be the velocity of the flight platform. a f is the Doppler frequency. prfIndicates the pulse repetition frequency, ΔR″ m,i (r,f a Let be the residual slant range error between channel m and the reference channel due to the cross-trajectory baseline and azimuth beam. The reconstruction filter H can be obtained using the above equation. -1 (f a ).

[0057] To further improve processing efficiency, ΔR″ can be calculated. m,i (r,f a The imaging error caused by spatial variation of distance, if the ambiguity component produced by this approximation is less than the system's azimuth ambiguity ratio (AASR) requirement within a certain range, is considered to be within this range. m,i (r,f a The distance variation is negligible and can be determined by ΔR″. m,i (r0,f a Let r0 represent the slant distance from the center of the target scene. At this point, the signal steering matrix P(r,f) a ) can be approximated as P′(f) a The transfer function of a linear system composed of distributed SAR can be further simplified to:

[0058] H′(f a )=B(f a )P′(f a (5)

[0059] Wherein, P′(f a ) is to add ΔR″ in equation (4) m,i (r,f a Replace ) with ΔR″ m,i (r0,f a P(r,f) a Let H' be a column vector. Using the above equation, the simplified reconstruction filter can be obtained as H''. -1 (f a ), using H′ -1 (f a This allows for distributed SAR spectrum recovery under the azimuth wide-beam assumption.

[0060] To further verify the effectiveness of the invention, an equivalent verification experiment was conducted in a certain area of ​​Shaanxi Province using airborne dual-channel SAR to image a farmland scene. Figure 2 The image shows the optical layout of the experimental scene. For the proposed algorithm, the signal processing flow of the dual-channel system is completely consistent with that of the multi-channel system; therefore, the proposed algorithm can be verified using the dual-channel system. In addition to farmland, the scene contains some houses with strong scattering. The system used is a dual-channel radar system with a hybrid baseline, employing a one-transmit, multiple-receiver operating mode. Specific operating parameters are shown in the table below.

[0061]

[0062] According to the parameters shown in the table above, the intervals of the equivalent phase centers between channels in the x, y, and z directions (azimuth, distance from ground, and altitude in the aircraft body coordinate system) are 1.85m, 0.10m, and 0.04m, respectively.

[0063] Figure 3 shows the imaging results after processing with the traditional spectrum reconstruction algorithm and the proposed spectrum reconstruction algorithm. As shown in Figure 3(a), the traditional transfer function method can reduce blur energy to some extent, but significant blur residue remains for strongly scattering targets such as target A and target B. Furthermore, compared to the narrow azimuth beam assumption, the blur components are restored to incorrect spectral positions after spectrum reconstruction, and range migration is not correctly corrected, resulting in severe defocusing in the imaging results. As shown in Figure 3(b), the blur components within the red box are virtually invisible in the imaging results after spectrum reconstruction using the proposed method, and the image quality is significantly improved.

[0064] After spectrum reconstruction using the traditional algorithm, Figures 4(a) and 4(b) show that target A has an ambiguity component in the 11360th azimuth cell, with an AASR of approximately -23.9 dB. Target B has an ambiguity component in the 9816th azimuth cell, with an AASR of approximately -18.5 dB. The ambiguity suppression is improved after spectrum reconstruction using the traditional method. However, SAR system design typically requires an AASR better than -25 dB, which the traditional method cannot meet. As shown in Figures 4(c) and 4(d), using the proposed algorithm, the ambiguity energy is suppressed to clutter levels, and the imaging quality is significantly improved.

[0065] In summary, this section fully validates the effectiveness of the proposed algorithm using measured data processing. The improved azimuth wide-beam distributed SAR spectrum reconstruction algorithm can effectively compensate for azimuth spatially varying channel mismatch and suppress ambiguity.

[0066] Of course, there are many other examples of this invention. Without departing from the spirit and essence of this invention, those skilled in the art can make various corresponding changes and modifications according to this invention, but these corresponding changes and modifications should all fall within the protection scope of the appended claims.

[0067] The contents not described in detail in this specification are existing technologies known to those skilled in the art.

[0068] The foregoing description illustrates and describes several preferred embodiments of the present invention. However, as previously stated, it should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the inventive concept described herein through the foregoing teachings or techniques or knowledge in related fields. Any modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

Claims

1. A method for azimuth wide-beam distributed SAR spectrum reconstruction, characterized in that... The steps of this method include: S1, perform range compression and range-directed registration on the raw echo data to obtain the range-directed registered data; S2, perform range spatial phase mismatch compensation on the range-registered data obtained in step S1 to obtain mismatch-compensated data; S3, perform deviation correction on the phase error estimation and compensation of the OSM channel of the mismatch-compensated data obtained in step S2 to obtain the error-compensated data; S4. Perform spectrum reconstruction or azimuth channel mismatch on the error-compensated data obtained in step S3 to obtain the unambiguous signal Doppler spectrum after azimuth channel mismatch or spectrum reconstruction. S5. Perform SAR imaging based on the unambiguous signal Doppler spectrum obtained from the azimuth channel mismatch or spectrum reconstruction in step S4 to obtain a SAR image. In step S2, range spatially variable phase mismatch compensation is performed on the range-registered data obtained in step S1 based on the inertial navigation data. In step S4, the method for performing spectrum reconstruction or azimuth channel mismatch on the error-compensated data is as follows: First, the imaging error is calculated when azimuth spatial variation is ignored, and the imaging error is judged. When the imaging error is less than or equal to a set threshold, the imaging error is ignored, and the spectrum is reconstructed on the error-compensated data using the transfer function method. When the imaging error is greater than the set threshold, the imaging error cannot be ignored, and the corresponding azimuth channel mismatch is compensated.

2. The azimuth wide-beam distributed SAR spectrum reconstruction method according to claim 1, characterized in that: In step S1, the original echo data is range compressed and range-oriented registered by matched filtering.

3. The azimuth wide-beam distributed SAR spectrum reconstruction method according to claim 1, characterized in that: The method for range-space-varying phase mismatch compensation is as follows: after calculating the baseline between channels using inertial navigation data, the corresponding range-space-varying phase mismatch is obtained, and the range-space-varying phase mismatch is compensated by conjugate multiplication in each channel to obtain the mismatch-compensated data.

4. The azimuth wide-beam distributed SAR spectrum reconstruction method according to claim 3, characterized in that: In step S4, the method for compensating for azimuth channel mismatch when the imaging error is not negligible is as follows: The transfer function of the linear system constructed from distributed SAR is: H(r,f a )=B(f a )P(r,f a ) (1) in: P(r,f a )=[p1(r,f a ),p2(r,f a ),...,p N (r,f a )] (3) Where M represents the number of channels, N represents the Doppler ambiguity number, M>N, diag{} represents a diagonal matrix with the elements within the brackets arranged diagonally, m=1,2,...,M, y m The y-axis coordinate of the equivalent phase center of the m-th channel receiving unit, v p Let λ be the velocity of the flight platform, λ be the carrier wavelength, r be the target slant range, and f be the velocity of the flight platform. a f is the Doppler frequency. prf f represents the pulse repetition frequency. dc Let f be the center frequency of the Doppler wave. a =f a1 +f dc +(M-1)f prf / 2,f a1 The range is (0, f) prf ],ΔR″ m,i (r,f a ) represents the residual slant range error of the two-dimensional spatial variation between channel m and the reference channel due to the cross-track baseline and azimuth beam.

5. The azimuth wide-beam distributed SAR spectrum reconstruction method according to claim 4, characterized in that: The reconstruction filter H is obtained according to equations (1)-(4). -1 (f a ).

6. A method for azimuth wide-beam distributed SAR spectrum reconstruction according to claim 4 or 5, characterized in that: When ΔR″ m,i (r,f a If the imaging error caused by distance spatial variation is within a set threshold range and the approximately generated blur component is less than the system's azimuth blur ratio (AASR) requirement, then within this range, ΔR″ m,i (r,f a The spatial variation of the distance is negligible, as determined by ΔR″. m,i (r0,f a Let r0 be the slant distance from the center of the target scene. Then, the signal steering matrix P(r,f) represents... a ) is approximately P′(f a The transfer function of a linear system composed of distributed SAR is: H′(f a )=B(f a )P′(f a ) (5) Wherein, P′(f a ) is to add ΔR″ in equation (4) m,i (r,f a Replace ) with ΔR″ m,i (r0,f a P(r,f) a ), which is a column vector.

7. The azimuth wide-beam distributed SAR spectrum reconstruction method according to claim 6, characterized in that: According to equation (5), the simplified reconstruction filter is H′. -1 (f a ), using H′ -1 (f a To achieve distributed SAR spectrum recovery under the azimuth wide beam assumption.

8. The azimuth wide-beam distributed SAR spectrum reconstruction method according to claim 1, characterized in that: In step S5, the method for performing SAR imaging based on the unambiguous signal Doppler spectrum after azimuth channel mismatch or spectrum reconstruction is as follows: the imaging process is performed using the range Doppler algorithm or the linear frequency modulation algorithm, and finally an unambiguous SAR image is obtained.

Images