Satellite-borne high-resolution wide swath SAR simultaneous multi-beam signal synthesis method

Abstract

The application relates to the field of spaceborne radars, and particularly discloses a signal synthesis method applied to a high-resolution wide-swath SAR system, which comprises the following steps: acquiring time sequence functions of inner calibration signals and echo receiving signals of multiple feed sources; the multiple feed sources emit the inner calibration signals through multiple beams; the multiple feed sources and the multiple beams are in one-to-one correspondence; the multiple beams are in one-to-one correspondence with multiple time domains; receiving signals through the multiple beams; each beam is used for receiving part of information of the receiving signals; performing discrete Fourier transform on the inner calibration signals, the time sequence functions and the receiving signals to solve a scattering coefficient matrix; performing inverse Fourier transform on the scattering coefficient matrix to obtain scene scattering signals; and the scene scattering signals contain multiple parts of information from the multiple beams respectively. The scheme provided by the application has high matching precision, can maximize the signal-to-noise ratio of the multiple beam overlapping parts, has the advantages of reducing sidelobes, and can meet the demand of high-resolution imaging.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of spaceborne radars, in particular to a spaceborne high-resolution wide-swath SAR simultaneous multi-beam signal synthesis method. BACKGROUND

[0002] Spaceborne high-resolution SAR imaging on the ground can realize all-time and all-weather limitations, filling the spaceborne high-resolution imaging detection blank. The spaceborne SAR radar system works in low orbit, and the theoretical azimuth resolution of the traditional strip SAR system is half of the size of the antenna azimuth direction, so reducing the size of the antenna in the azimuth direction can improve the azimuth resolution. At the same time, a wide beam is obtained by reducing the size of the antenna in the elevation direction to improve the distance mapping bandwidth. However, the size of the antenna in the azimuth direction and the distance direction cannot be infinitely reduced, and in order to meet certain performance indicators, there is a minimum limit for the area of the SAR antenna. A SAR smaller than the antenna area will not work normally.

[0003] In order to obtain higher resolution, the azimuth resolution of the sliding spotlight mode is adopted, the imaging range in the azimuth direction is greater than that in the spotlight mode, and the imaging resolution is better than that in the strip mode. The mosaic mode is a relatively effective high-resolution wide-swath imaging mode, but compared with the sliding spotlight mode, the resolution in the azimuth direction is sacrificed, and it is difficult for the imaging resolution to reach more than 0.5 m. When the resolution reaches more than 0.2 m, the observation range in the distance direction and the azimuth direction reaches more than 10 km, at this time, the required quality factor reaches more than 50. An effective solution is to use distance simultaneous multi-beam technology. However, since different beams emit the same linear frequency modulation pulse signal, the amplitude and phase errors between multiple beams need to be controlled so that the amplitude error is within 0.3 dB and the phase error is within 10°. It is difficult to achieve such high errors in the radar system, and the errors must be estimated and compensated in signal processing. SUMMARY

[0004] The application provides a signal synthesis method and a radar device, which aims to improve the signal quality in the overlapping area, match high precision, maximize the signal-to-noise ratio of the overlapping part of multiple beams, and reduce the sidelobe advantage, which can meet the demand of sub-meter high-resolution imaging.

[0005] In a first aspect, a signal synthesis method is provided, which is applied to a high-resolution wide-swath SAR system, comprising:

[0006] obtaining an inner calibration signal s i (t) of multiple feeds and a time sequence function u i (t) of echo receiving signals, i represents the number of feeds, and the time sequence function u i (t) satisfies:

[0007] T i is the start time of the corresponding beam echo of the plurality of feed sources, T Ri represents the width of the corresponding beam echo of the i-th beam;

[0008] The plurality of feed sources transmits an internal calibration signal s i (t), the plurality of feed sources and the plurality of beams are one-to-one corresponding, and the plurality of beams correspond to a plurality of time domains one-to-one;

[0009] The plurality of beams receive a signal x(t), and each beam is used to receive a part of information of the received signal x(t);

[0010] According to the internal calibration signal s i (t), the time sequence function u i (t) and the received signal x(t), a discrete Fourier transform is performed to solve a scattering coefficient matrix H(f);

[0011] Inverse Fourier transform is performed on the scattering coefficient matrix H(f) to obtain a scene scattering signal h(t), which contains a plurality of parts of information from the plurality of beams respectively.

[0012] Compared with the prior art, the scheme provided by the application has at least the following beneficial technical effects:

[0013] The patent analyzes a multi-beam receiving signal model, uses the characteristics that the calibration signal of the multi-beam transmitting signal and the time window of the received signal are known, gives a joint pulse compression method, calculates the echo signal in the receiving window, and improves the signal quality in the overlapping area. The multi-beam matched filter function construction method has the advantages of high matching accuracy, can maximize the signal-to-noise ratio of the overlapping part of the plurality of beams, and reduce the sidelobe, and can meet the demand of sub-meter high-resolution imaging. The scene signal recovered by the method is consistent with the preset signal, and the method effectively solves the problem of segmented echo recovery scene signal, and has important significance for the system development of high-resolution wide-width SAR.

[0014] In combination with the first aspect, in some implementations of the first aspect, the method specifically comprises:

[0015] Sampling the received signal to obtain information

[0016] Performing N r point discrete Fourier transform on the transmitting signal s i (n), the time sequence function u i (n) and the received signal x(n) to obtain X[k] = ΓH[k], is a diagonal matrix, the diagonal elements of which are composed of S i [k], S i [k] is composed of the transmitted signal s i (n) is obtained by discrete Fourier transform, is a matrix constructed by U i [k], U i [k] is composed of the time series function u i (n) is obtained by discrete Fourier transform,

[0017] X[k] = ΓH[k] is converted into H(f) = Γ -1 X(f), and H(f) = Γ -1 X(f) is solved to obtain the scattering coefficient matrix H(f).

[0018] With reference to the first aspect, in some implementations of the first aspect, solving H(f) = Γ -1 X(f) includes:

[0019] singular value decomposition is performed on Γ to obtain Γ = UΛV H , where Λ = diag(γ) is a diagonal matrix composed of singular values, γ = [γ1 γ2 … γ P 0 … 0] is used to represent the singular values, P is the number of non-zero singular values, and U and V are corresponding unitary matrices.

[0020] The truncated pseudo-matrix of the matrix Γ is calculated as wherein,

[0021] Γ -1 is replaced by , and the scattering coefficient matrix H(f) is calculated according to .

[0022] Solving the scattering coefficient matrix H(f) by singular value truncation is conducive to reducing the processing amount required for solving the scattering coefficient matrix H(f).

[0023] With reference to the first aspect, in some implementations of the first aspect, the sampling rate is 0.05-3 GHz.

[0024] With reference to the first aspect, in some implementations of the first aspect, the received signal x t is used to represent a target image, the plurality of beams are used to receive a plurality of different image blocks of the target image, and the plurality of beams correspond one-to-one to the plurality of image blocks; the method further includes:

[0025] According to the scene scattering signal h(t), the target image is imaged.

[0026] The obtained scene scattering signal h(t) can contain information of the whole image. The image splicing processing amount can be reduced without splicing the signals received by multiple beams.

[0027] With reference to the first aspect, in some implementations of the first aspect, the method is applied to a scene with time domain overlap between echo windows of two adjacent beams.

[0028] With reference to the first aspect, in some implementations of the first aspect, the number of the plurality of feeds is greater than or equal to 4.

[0029] The second aspect provides a radar device for performing the method in any of the implementations of the first aspect. BRIEF DESCRIPTION OF DRAWINGS

[0030] Figure 1 A schematic flowchart of a signal synthesis method provided by an embodiment of the present application.

[0031] Figure 2 A schematic diagram of a signal synthesis method provided by an embodiment of the present application.

[0032] Figure 3 A schematic flowchart of a radar device provided by an embodiment of the present application.

[0033] Figure 4 A result map after matching of signal 1.

[0034] Figure 5 A result map after matching of signal 2.

[0035] Figure 6 A result map after matching of signal 3.

[0036] Figure 7 A result map after matching of signal 4.

[0037] Figure 8 A comparison map of a signal recovered by a signal synthesis method provided by an embodiment of the present application and a preset signal. DETAILED DESCRIPTION

[0038] The present application will be further described in detail below with reference to the drawings and specific embodiments.

[0039] Figure 1 A schematic flowchart of a signal synthesis method provided by an embodiment of the present application.

[0040] Step 1: Obtain internal calibration signals s i (t) of a plurality of feeds, where i represents the number of the feed. The internal calibration signals s i(t) can have the characteristics of consistent pulse width and consistent repetition frequency.

[0041] Step 2: Obtain the time series function u i (t)

[0042] W i = T i_end -T i_start .

[0043] High-diversity wide swath SAR transmits multiple signals simultaneously through multiple feeds. T i is the start time of the beam echo corresponding to the multiple feeds, determined by the wave position of the system. T Ri represents the wide swath echo time corresponding to the i-th beam. T i_start , T i_end are the start time and end time of the beam echo corresponding to the multiple feeds, respectively, determined by the wave position of the system. C represents the speed of light. T p is the pulse width. The time series function u i (t) can be determined by measuring the geometry of the star-ground relationship and the range of the echo signal according to the multiple feed signals.

[0044] Step 3: Multiple feeds transmit internal calibration signals s i (t) (s i (t) can represent the transmitted signal), one-to-one corresponding to multiple feeds and multiple beams, and receive signals x(t) through multiple beams, the received signal x(t) satisfies: h(t) is the scattering coefficient vector of the scene, representing the integration of the equidistance ring of the wide swath surface target. h(t) can be a parameter to be solved. The scene scattering signal h(t) can be used to generate an image. represents a circular convolution operation.

[0045] In one embodiment, in some embodiments, the number of multiple feeds can be 4. As Figure 2 shown, for a certain pulse, the signals transmitted and received by the four feeds are

[0046]

[0047] can be expressed as

[0048]

[0049] wherein

[0050] The received signal x(t) can be sampled. The sampling rate is F s , obtaining N i = T i ·Fs , N Ri = T Ri · F s , N p = T p · F s , N r = T r · F s , T r may be the pulse width pulse repetition period. The above equation becomes;

[0051]

[0052] x(n) can be used to represent x(t). s i (n) can be used to represent s i (t). u i (n) can be used to represent u i (t).

[0053] Step 4: Perform Fourier transform on the transmit signal s i (t), the time series function u i (t), and the received signal x(t). In some embodiments, Fourier transform can be performed on the transmit signal s i (n), the time series function u i (n), and the received signal x(n).

[0054] Specifically, performing N r point Discrete Fourier Transform (DFT) on the above equation gives

[0055]

[0056] where is a diagonal matrix, whose diagonal elements are composed of S i [k], S i [k], U i [k], H[k], X[k] are all N r × 1 column vectors, is an N i × N i matrix constructed from S r [k], U r [k]. Then the circular convolution can be represented by matrixization, i.e.

[0057]

[0058] where,

[0059] Then H(f) = Γ -1 (f)X(f), where Γ(f) is determined by S i (f), U i (f) can be determined by the calibration signal and the geometric relationship, and can be regarded as a known quantity.

[0060] In theory, the expected signal H(f) can be directly solved. In practice, Γ -1 (f) is an underdetermined matrix. Embodiments of the present application use a truncated approximation method to achieve.

[0061] Step 5: Singular value decomposition (SVD) analysis is performed on the transformation matrix Γ -1 (f), the eigenvalues γ1…γ P are calculated, and the order P is determined. The order P is the number of non-zero singular values.

[0062] Specifically, to simplify the formula, Γ is used to represent Γ(f) in the following, and singular value decomposition SVD is performed on Γ to obtain Γ = UΛV H .

[0063] Where Λ = diag(γ) is a diagonal matrix composed of singular values, γ = [γ1 γ2…γ P 0…0] is used to represent the singular values, P is the number of non-zero singular values, and U and V are corresponding unitary matrices.

[0064] Step 6: Calculate the truncated approximation matrix of the transformation matrix.

[0065] Specifically, the truncated approximation of the matrix Γ is

[0066] Where,

[0067] Step 7: Calculate the radar echo signal scattering coefficient matrix H(f) inversely.

[0068] Then use to replace Γ -1 to calculate

[0069] Step 8: Perform inverse fast Fourier transform (IFFT) on the scattering coefficient matrix H(f) to obtain the scene scattering signal h(t). The scene scattering signal can be used as input for subsequent imaging.

[0070] In some possible scenarios, the radar device can transmit a signal s i(t), multiple feed sources and multiple beams correspond one-to-one. Radar equipment can receive signals x(t) through multiple beams, and the received signal x(t) can represent a target image. Multiple beams can transmit different image patch information of the target image, and multiple beams can correspond one-to-one with multiple image patch information. For example... Figure 2 As shown, signals received by multiple beams intersect in the time domain. Discarding intersecting signals due to interference can reduce the integrity of information reception. However, improper handling of intersecting signals results in poor signal processing quality. The scene scattering signal h(t) obtained by the method provided in this application can contain information about the intersecting signals, and the signal processing quality is excellent. The target image can then be obtained from the scene scattering signal h(t). In other words, the obtained scene scattering signal h(t) can contain information about the entire image. In one embodiment, it is not necessary to stitch together signals received from multiple beams, reducing the amount of image stitching processing.

[0071] This application also provides a radar device, such as... Figure 3 As shown, radar equipment may include a digital-to-analog converter (DAC) to generate one signal, which is then divided by a power divider and transmitted to multiple transmission channels to form a high-power signal. This signal is then simultaneously transmitted via multiple transmit feeds as a linear frequency modulated signal. The electromagnetic wave signal is reflected by an antenna reflector to reach the target to be observed. After being backscattered by the target, the electromagnetic wave is focused by the antenna reflector and simultaneously received from multiple feeds. It undergoes low-noise amplification and frequency conversion through multiple channels, and finally, after analog-to-digital conversion (AD) sampling, a digital domain signal is obtained.

[0072] Example 1

[0073] The simulation test settings are as follows:

[0074] Serial number Item Value 1. Sampling rate 50MHz 2. Signal 1 Positive chirp LFM signal with a bandwidth of 20MHz 3. Signal 2 Negative chirp LFM signal with a bandwidth of 20MHz 4. Signal 1 Windowed positive chirp LFM signal with a bandwidth of 20MHz 5. Signal 2 Windowed negative chirp LFM signal with a bandwidth of 20MHz 6. Pulse width 10us 7. Scene echo duration 34us

[0075] The figure below shows the scene signals and corresponding echo window simulation settings for the four signals, with a 2µs overlap between each segment. Figures 4-7 The results show the superimposed echoes of the four signal segments and the matched filtering results with the transmitted signal. It can be seen that the front-end signal result is consistent with the original signal, while the back-end signal is a mismatched signal. Figure 8 To obtain the scene signal using the proposed algorithm, it can be seen that the scene signal recovered by the algorithm is consistent with the preset signal. The algorithm effectively solves the problem of recovering scene signals from segmented echoes, which is of great significance for the development of high-resolution wide-swath SAR systems.

[0076] Although the present application is disclosed with reference to the preferred embodiments above, it is not intended to limit the present application, and any person skilled in the art can make possible variations and modifications without departing from the spirit and scope of the present application. Therefore, the scope of protection of the present application should be defined by the scope of the claims.

Claims

1. A signal synthesis method, characterized in that, The method is applied to a high-resolution wide-swath SAR system, including: Acquire internal calibration signals s from multiple feed sources i (t) and the time series function u of the echo received signal i (t), where i represents the feed source number, and the time series function u i (t) satisfies: T i T represents the start time of the beam echo corresponding to the multiple feed sources. Ri This represents the swath width echo time corresponding to the i-th beam; The multiple feed sources transmit internal calibration signals s through multiple beams. i (t), the plurality of feed sources and the plurality of beams correspond one-to-one, and the plurality of beams correspond one-to-one with the plurality of time domains; The signal x(t) is received through the plurality of beams, each beam being used to receive a portion of the information of the received signal x(t); According to the internal calibration signal s i (t), the time series function u i Perform a discrete Fourier transform on the received signal x(t) and x(t) to solve for the scattering coefficient matrix H(f); Performing an inverse Fourier transform on the scattering coefficient matrix H(f) yields the scene scattering signal h(t), which contains multiple portions of information from the plurality of beams; wherein, For the received signal Sampling to obtain information x(n) is used to represent x(t), s i (n) is used to represent s i (t), u i (n) is used to represent u i (t); For the transmitted signal s i (n), Time series function u i (n), Receive signal x(n) and execute N r Point-based discrete Fourier transform yields X[k] = ΓH[k]. It is a diagonal matrix, and the diagonal elements are determined by S. i [k] constitutes, S i [k] is the transmitted signal s i (n) is obtained through discrete Fourier transform. For U i The matrix constructed by [k], U i [k] is derived from the time series function u i (n) is obtained through discrete Fourier transform. Convert X[k]=ΓH[k] to H(f)=Γ -1 X(f), and for H(f)=Γ -1 X(f) is solved to obtain the scattering coefficient matrix H(f).

2. The method according to claim 1, characterized in that, The pair H(f)=Γ -1 Solving for X(f) includes: Singular value decomposition of Γ yields Γ = UΛV H , where Λ=diag(γ) is a diagonal matrix composed of singular values, γ=[γ1 γ2 … γ P 0 … 0] is used to represent singular values, P is the number of non-zero singular values, and U and V are the corresponding unitary matrices; The truncated quasi-matrix of matrix Γ is calculated as follows: in, use Replace Γ -1 and according to The scattering coefficient matrix H(f) is calculated.

3. The method according to claim 1, characterized in that, The sampling rate is 0.05–3 GHz.

4. The method according to any one of claims 1 to 3, characterized in that, The received signal x(t) is used to represent the target image, and the plurality of beams are used to receive different image patches of the target image, with each beam corresponding one-to-one with the image patches; the method further includes: The target image is obtained by imaging based on the scene scattering signal h(t).

5. The method according to any one of claims 1 to 3, characterized in that, The method is applicable to scenarios where there is time-domain overlap between the echo windows of two adjacent beams.

6. The method according to any one of claims 1 to 3, characterized in that, The number of the plurality of feed sources is greater than or equal to 4.

7. A radar device, characterized in that, The radar device is used to perform the method as described in any one of claims 1 to 6.

Images