A distributed small satellite spotlight MIMO-SAR ultra-high resolution imaging method
By demarcating molecular aperture signals in distributed small satellite beam MIMO-SAR system and using improved time domain bandwidth synthesis method and image fusion algorithm, the problem of Doppler aliasing and excessive data volume is solved, and the problem of echo signal Doppler aliasing and excessive data volume is realized. Efficient ultra-high resolution imaging simplifies operational processes and reduces computational volume.
Patent Information
- Application Number
- CN202011586068.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2020-12-20
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2040-12-20
AI Technical Summary
When the distributed small satellite beam MIMO-SAR system realizes high-resolution wide mapping band imaging, it faces the problems of Doppler aliasing and excessive data volume of echo signals. The traditional bandwidth synthesis method is cumbersome and inefficient, resulting in imaging. The processing system has a lot of calculations.
By dividing the pore size signals, azimuth blur is eliminated using airspace filtering technology, and an improved time domain bandwidth synthesis method and sub-aperture image fusion algorithm are used to reconstruct distance-to-large bandwidth signals and azimuth high-resolution images, simplifying the operation process and reducing the calculation amount.
Effectively reduce the ambiguity of the processed signal and the calculation amount of the imaging system, improve the band synthesis efficiency, realize ultra-high resolution imaging of distributed small satellite beam MIMO-SAR, and reduce the on-site storage pressure and imaging processing. load.
Smart Images

Figure CN114646958B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of small satellite SAR, and in particular to an ultra-high resolution imaging method for a distributed small satellite spotlight MIMO-SAR working mode. Background Art
[0002] MIMO-SAR overcomes the contradiction between high resolution and wide mapping swath of single-channel SAR, and is the key development direction of spaceborne SAR systems in the future. Phased array antenna technology is a major way to achieve the spaceborne MIMO-SAR working mode, but it greatly increases the development cost of radars. Distributed small satellites have become another way to achieve the spaceborne MIMO-SAR working mode in the future due to their advantages such as flexible launch, short development cycle and low cost. Distributed small satellites can complete the spaceborne beamforming MIMO-SAR configuration by controlling the antenna beam pointing, and multiple antennas simultaneously transmit and receive signals in multiple frequency bands, thus achieving high-resolution wide mapping swath imaging.
[0003] The beamforming mode is the main means for SAR to obtain high-resolution target images. In this mode, the Doppler bandwidth of the full-aperture echo signal is large. In order to avoid Doppler aliasing of the echo signal, the system pulse repetition rate (PRF) is generally required to be greater than this Doppler bandwidth, but this will greatly limit the mapping bandwidth of the observation scene. In order to obtain a large mapping bandwidth, two types of MIMO-SAR algorithms are widely used: range ambiguity suppression algorithm and azimuth deambiguation algorithm. The range ambiguity suppression algorithm is often implemented through azimuth phase coding (APC). However, this type of algorithm requires that the system PRF is greater than the Doppler bandwidth of the echo signal, which will undoubtedly increase the amount of data collected by the radar receiver and cause excessive pressure on the satellite storage. Therefore, spaceborne beamforming MIMO-SAR usually hopes to work in a low PRF mode and eliminate azimuth ambiguity by using an azimuth deambiguation algorithm. The invention uses spatial degrees of freedom to perform spatial filtering on the full-aperture signal to eliminate the azimuth ambiguity of the signal. However, since the spatial degrees of freedom are limited by the number of satellites, when the resolution is improved and the azimuth ambiguity is serious, the number of satellites needs to be increased to obtain sufficient spatial degrees of freedom, which will undoubtedly lead to a sharp increase in costs. The invention performs deambiguation processing on the sub-aperture signal, which greatly reduces the ambiguity of the processed signal, thereby ensuring sufficient spatial degrees of freedom. However, this type of algorithm needs to splice the full-aperture signal before imaging, which causes a surge in the amount of data and greatly increases the computational complexity of the imaging processing system.
[0004] In a high-resolution SAR system, it is necessary to transmit a wideband signal to obtain high resolution in the range direction. However, ultra-wideband signals have high requirements for the transmitting hardware equipment and are difficult to generate directly by hardware. One way to solve this problem is for the radar to transmit a narrowband stepped frequency signal, and then process the radar echo through bandwidth synthesis technology to obtain an ultra-wideband signal. The bandwidth synthesis technology of radar signals is mainly divided into two categories: time-domain bandwidth synthesis (TBS) and frequency-domain bandwidth synthesis (FBS). The FBS method is simple to operate, but it requires that the sub-band center of the signal is exactly separated from the center of the complete frequency band by an integer number of frequency points, otherwise it cannot be accurately synthesized. The TBS method has high synthesis accuracy, but its operation process is relatively cumbersome, the amount of calculation is large, and the efficiency is low.
[0005] Based on the above problems, the present invention proposes an ultra-high resolution imaging method for distributed small satellite beam MIMO-SAR, which obtains high-resolution distance in the distance dimension by transmitting stepped frequency signals, and obtains high-resolution azimuth images in the azimuth dimension by sub-aperture image fusion. First, the spatial domain filtering technology is used to remove the azimuth ambiguity of the sub-aperture signal, and then an improved TBS method is proposed to realize the synthesis of stepped frequency signals. Finally, the sub-aperture image fusion algorithm is used to complete the imaging and image fusion of the sub-aperture signal, thereby obtaining a high-resolution image corresponding to the full aperture. The combination of sub-aperture division technology and sub-aperture image fusion algorithm effectively reduces the ambiguity of the processed signal and the computational complexity of the imaging system. The improved TBS method simplifies the operation process of the traditional TBS and improves the efficiency of frequency band synthesis. Simulation experiments verify the feasibility and effectiveness of this method. Summary of the invention
[0006] The purpose of the present invention is to propose an ultra-high resolution imaging method for distributed small satellite spotlight MIMO-SAR, which obtains high range resolution in the range dimension by transmitting stepped frequency signals, and obtains high azimuth resolution images in the azimuth dimension by sub-aperture image fusion, so as to obtain high range resolution.
[0007] In order to achieve the above object, the technical solution adopted by the present invention is:
[0008] A distributed small satellite spotlight MIMO-SAR ultra-high resolution imaging method, characterized by comprising the following steps:
[0009] Step 1: Divide the sub-aperture and separate the sub-band signals
[0010] (1) Divide the full-aperture signal into K sub-aperture signals.
[0011] (2) Separate the sub-aperture signal into sub-band signals, and separate the signal received by each satellite into Q sub-band signals, where Q is the number of satellites transmitting the signal.
[0012] (3) Perform Doppler center compensation on the separated sub-aperture sub-band signals to make their Doppler centers zero.
[0013] Step 2: Reconstruct the position without ambiguity signal
[0014] (1) After Doppler center compensation, the sub-aperture signal is subjected to azimuth FFT processing.
[0015] (2) A spatial filtering weight vector function W(m) is constructed according to the spatial degrees of freedom of the signal in azimuth. The sub-aperture signal is filtered using W(m) to obtain an azimuthally unambiguous signal.
[0016] Step 3: Reconstruct the distance-oriented large bandwidth signal
[0017] (1) First, the sub-band signals without azimuth ambiguity are transformed into the dual-frequency domain by range FFT processing, and then the quadratic term of the range phase is compensated.
[0018] (2) Use the improved TBS method to complete bandwidth splicing:
[0019] a) Perform range IFFT processing on each sub-band signal to transform it into the range time domain, perform range frequency shift in the time domain, and then perform range FFT operation.
[0020] b) The sub-band signals are coherently accumulated to obtain a complete range-frequency band signal. After the frequency band synthesis, the bandwidth of the signal in the range direction is expanded from B to QB, where B is the bandwidth of the sub-band signal.
[0021] Step 4: Range motion compensation and range compression: Use CSA to perform range motion compensation and range pulse compression on the two-dimensional reconstructed signal of the sub-aperture.
[0022] Step 5: Azimuth Focus
[0023] (1) After the range pulse compression, the azimuth phase of the subaperture signal is a hyperbolic phase, which is converted into a quadratic phase in the azimuth frequency domain.
[0024] (2) Perform IFFT processing on the signal in the azimuth direction, and then perform Dechirp operation on the signal in the azimuth time domain.
[0025] (3) Correct the signal Doppler center that has changed due to the Dechirp operation so that the signal Doppler center is zero.
[0026] (4) Perform FFT operation on the signal in azimuth direction to obtain Azimuth low-resolution image of the focus area.
[0027] Step 6: Sub-aperture image fusion
[0028] (1) The azimuth phase of the sub-aperture image is corrected so that the phase of the sub-aperture image is zero at the focusing frequency point, and is linearly related to the center moment of the Kth sub-aperture at the non-focusing point.
[0029] (2) The sub-aperture images are coherently superimposed to obtain an azimuth full-resolution image.
[0030] Compared with the prior art, the beneficial effects of the present invention are as follows:
[0031] (1) The present invention improves the traditional TBS. The traditional TBS method has four steps: time domain frequency shift, phase correction, time shift, and frequency band accumulation. The two steps of phase correction and time shift are to correct the frequency shift of the secondary phase, and these two steps are cumbersome to calculate. The improved TBS method of the present invention compensates for the secondary phase of the distance before the frequency shift, so there is no need for phase correction and frequency shift operations, reducing the amount of calculation.
[0032] (2) The present invention designs a set of processing procedures for distributed small satellite beam MIMO-SAR. The traditional MIMO-SAR processing procedures are either for full-aperture signals or require splicing of full-aperture signals before imaging. The present invention processes sub-aperture signals and uses spatial domain filtering technology and an improved TBS method to restore sub-aperture signals with unambiguous azimuth and large bandwidth. Then, the sub-aperture signal is directly imaged: the sub-aperture signal is directly imaged using a sub-aperture image fusion imaging method to obtain multiple sub-aperture images, and then the sub-aperture images are coherently fused to obtain a high-resolution image corresponding to the full aperture. In this way, there is no need to splice the full-aperture signal for imaging, and parallel processing is possible. In addition, the amount of sub-aperture signal is much smaller than that of full-aperture signal, and the processor load is much smaller.
[0033] (3) The present invention adopts a distributed small satellite model. Compared with the single-station phased array model, the distributed small satellite can store data in a distributed manner, reducing the storage pressure on the satellite. BRIEF DESCRIPTION OF THE DRAWINGS
[0034] Figure 1 The present invention discloses a flow chart of a distributed small satellite spotlight MIMO-SAR ultra-high resolution imaging method.
[0035] Figure 2 The model of the satellite-borne three-transmitter and three-receiver system is shown.
[0036] Figure 3 The results of frequency band synthesis for a single point target: (a) is the bandwidth of the sub-band signal; (b) is the bandwidth of the signal after frequency band synthesis.
[0037] Figure 4 The distribution of point targets is shown: there is a 3×3 evenly distributed dot matrix on the ground, and the point spacing in the range and azimuth directions is 1km.
[0038] Figure 5 This is the distance profile of P1 before and after frequency band synthesis.
[0039] Figure 6 The change of the azimuth resolution of the point target P1 during the image fusion process is shown.
[0040] Figure 7 The contour maps of the final images of three point targets P1, P2, and P3 are shown: (a) is the contour map of point target P1; (b) is the contour map of point target P2; (c) is the contour map of point target P3.
[0041] Figure 8 This is the original image used for simulation.
[0042] Fig. 9 This is the imaging result of the surface target. DETAILED DESCRIPTION
[0043] Step 1: Divide the sub-aperture and separate the sub-band signals
[0044] (1) Divide the full-aperture signal into K sub-aperture signals.
[0045] (2) Separate the sub-aperture signal into sub-band signals, and separate the signal received by each satellite into Q sub-band signals, where Q is the number of satellites transmitting the signal.
[0046] (3) Perform Doppler center compensation on the separated sub-aperture sub-band signals to make their Doppler centers zero.
[0047] Step 2: Reconstruct the position without ambiguity signal
[0048] (1) After Doppler center compensation, the sub-aperture signal is subjected to azimuth FFT processing.
[0049] (2) A spatial filtering weight vector function W(m) is constructed according to the spatial degrees of freedom of the signal in azimuth. The sub-aperture signal is filtered using W(m) to obtain an azimuthally unambiguous signal.
[0050] Step 3: Reconstruct the distance-oriented large bandwidth signal
[0051] (1) First, the sub-band signals without azimuth ambiguity are transformed into the dual-frequency domain by range FFT processing, and then the quadratic term of the range phase is compensated.
[0052] (2) Perform range IFFT processing on each sub-band signal to transform it into the range time domain, perform range frequency shift in the time domain, and then perform range FFT operation.
[0053] (3) The sub-band signals are coherently added to obtain a complete range-band signal. After the frequency band synthesis, the bandwidth of the signal in the range direction is expanded from B to QB, where B is the bandwidth of the sub-band signal.
[0054] Step 4: Range motion compensation and range compression: Use CSA to perform range motion compensation and range pulse compression on the two-dimensional reconstructed signal of the sub-aperture.
[0055] Step 5: Azimuth Focus
[0056] (1) After the range pulse compression, the azimuth phase of the subaperture signal is a hyperbolic phase, which is converted into a quadratic phase in the azimuth frequency domain.
[0057] (2) Perform IFFT processing on the signal in the azimuth direction, and then perform Dechirp operation on the signal in the azimuth time domain.
[0058] (3) Correct the signal Doppler center that has changed due to the Dechirp operation so that the signal Doppler center is zero.
[0059] (4) Perform FFT operation on the signal in azimuth direction to obtain Azimuth low-resolution image of the focus area.
[0060] Step 6: Sub-aperture image fusion
[0061] (1) The azimuth phase of the sub-aperture image is corrected so that the phase of the sub-aperture image is zero at the focusing frequency point, and is linearly related to the center moment of the Kth sub-aperture at the non-focusing point.
[0062] (2) The sub-aperture images are coherently superimposed to obtain an azimuth full-resolution image.
[0063] The effectiveness of the present invention is further illustrated by target simulation experiments as follows.
[0064] Simulation experiment:
[0065] (1) Simulation conditions:
[0066] The simulation experiment uses Figure 2 The satellite-borne three-transmitter and three-receiver system model is shown in FIG. Each satellite transmits a linear frequency modulation signal with a different carrier frequency and receives signals from three satellites. The simulation parameters are shown in Table 1.
[0067] Table 1 Simulation parameters of the three-transmitter-three-receiver system
[0068]
[0069] (2) Simulation content:
[0070] In order to better show the signal spectrum, a single point target is first used to perform a frequency band synthesis simulation experiment. The single point target is located in the center of the scene. Figure 3 Band synthesis results for a single point target: The three sub-band signals can be well synthesized into a large bandwidth signal through the improved TBS method.
[0071] After verifying the effectiveness of the improved TBS method, imaging simulation of the dot matrix target was carried out. Figure 4 The distribution of point targets is shown: there is a 3×3 evenly distributed dot matrix on the ground, and the point spacing in the range and azimuth directions is 1km. Figure 4 The coordinates of the three points marked are P1(-1km, R ε -1km)、P2(0,R s )、P3(1km,R s +1km), R s is the shortest slant distance to the center of the scene.
[0072] The simulation results of point targets are as follows: Figure 5-7 shown. Figure 5 is the distance profile of P1 before and after band synthesis. Figure 5 It can be seen that after the frequency band synthesis, the image obtains high distance resolution. Figure 7 The figure shows the change of the azimuth resolution of the point target P1 during the image fusion process. Figure 6 It can be seen that the azimuth resolution of the fused image gradually increases with the increase in the number of sub-aperture images.
[0073] Figure 7 The contour map of the final imaging of three point targets P1, P2, and P3 is shown. Table 2 analyzes the peak sidelobe ratio and integrated sidelobe ratio of the imaging results of these three point targets. Figure 7 Table 2 and Table 3 both show that the imaging performance of the method proposed in the present invention is good.
[0074] Table 2 PSLR and ISLR of point targets
[0075]
[0076] (3) Simulation analysis:
[0077] Since there is no measured data of space-borne MIMO-SAR, the present invention uses a measured airborne SAR image as a surface target for echo simulation. The target pixels are arranged on the slant range plane, and the simulation parameters are shown in Table 1. The original image used for simulation is as follows Figure 8 shown.
[0078] According to the processing algorithm and process proposed by the present invention, the imaging result of the simulated surface target is as follows: Fig. 9 shown. Fig. 9The imaging results shown are basically the same as the original Figure 1 This verifies the effectiveness of the algorithm proposed in this invention.
Claims
1. A distributed small satellite spotlight MIMO-SAR ultra-high resolution imaging algorithm, characterized in that: The following steps are involved: Step 1: Divide the sub-aperture and separate the sub-band signals (1) Divide the full aperture signal into K sub-aperture signals; (2) Separate the sub-aperture signal into sub-band signals, and separate the signal received by each satellite into Q sub-band signals, where Q is the number of satellites transmitting the signal; (3) performing Doppler center compensation on the separated sub-aperture sub-band signals to make their Doppler centers zero; Step 2: Reconstruct the position without ambiguity signal (1) After Doppler center compensation, perform azimuth FFT processing on the sub-aperture signal; (2) constructing a spatial filtering weight vector function W(m) according to the spatial degrees of freedom of the signal in azimuth, and using W(m) to filter the sub-aperture signal to obtain an azimuthally unambiguous signal; Step 3: Reconstruct the distance-oriented large bandwidth signal (1) First, perform range FFT processing on each sub-band signal with unambiguous azimuth to transform it into dual frequency domain, and then compensate for the quadratic term of range phase; (2) Use improved time-domain bandwidth synthesis (TBS) to complete bandwidth synthesis; Step 4: Range motion compensation and range compression: Use CSA to perform range motion compensation and range pulse compression on the two-dimensional reconstructed signal of the sub-aperture; Step 5: Azimuth Focus (1) After the range pulse compression, the azimuth phase of the subaperture signal is a hyperbolic phase, which is converted into a quadratic phase in the azimuth frequency domain; (2) Perform IFFT processing on the signal in the azimuth direction, and then perform Dechirp operation on the signal in the azimuth time domain; (3) Correct the signal Doppler center that has changed due to the Dechirp operation so that the signal Doppler center is zero; (4) Perform FFT operation on the signal in azimuth direction to obtain Azimuth low-resolution images of the domain focus; Step 6: Sub-aperture image fusion (1) Correcting the azimuth phase of the sub-aperture image so that the phase of the sub-aperture image is zero at the focusing frequency point, and is linearly related to the center moment of the Kth sub-aperture at the non-focusing point; (2) The sub-aperture images are coherently superimposed to obtain an azimuth full-resolution image.
