A Decoupled Hybrid Correlation Algorithm for High-Squint Spaceborne SAR Data Imaging

Fig. 1. Geometry model of the high-squint spaceborne SAR.

A Decoupled Hybrid Correlation Algorithm for High-Squint Spaceborne SAR Data Imaging

Y. Guo, P. Wang, Z. Men, T. He, T. Qiu and J. Chen, "A Decoupled Hybrid Correlation Algorithm for High-Squint Spaceborne SAR Data Imaging," in IEEE Transactions on Geoscience and Remote Sensing, vol. 62, pp. 1-18, 2024, Art no. 5218018, doi: 10.1109/TGRS.2024.3430294.

Abstract: High-resolution and high-squint spaceborne synthetic aperture radar (SAR) system has excellent earth observation performance. However, high-squint spaceborne SAR data is more challenging to process than the general broadside counterpart because of the severe range-azimuth coupling (RAC). Whereas the classic linear range walk correction (LRWC) method can handle this problem, it is performed in azimuth time-domain and seriously constrains azimuth swath width. 

In this article, a decoupled hybrid correlation Algorithm (DHCA), combining the range-azimuth decoupling (RAD) in 2-D frequency domain with the modified hybrid correlation (MHC), is proposed to handle sliding spotlight spaceborne SAR data of high-squint angle case. The decoupled hybrid correlation (DHC) is the main body of the proposed algorithm, and it starts with RAD filtering in 2-D frequency domain, which is designed to eliminate the majority of the range cell migration (RCM) and RAC caused by high-squint angles. A nonlinear chirp scaling (NLCS) in range direction is subsequently performed to equalize the variant range chirp rate caused by residual RCM and RAC. Coarse range compression can then be realized uniformly by the range-matched filtering. The NLCS operation and range coarse compression ensure that the range-Doppler (RD) signals of all targets within the whole scene occupy very narrow range cell scopes. 

By making full use of the principle of stationary point (POSP) to signals, the range compression position and range frequency mapping relationship after NLCS can be derived. The MHC can therefore be fulfilled by extracting the RD signals along the residual RCM and constructing the reference correlation function after the range NLCS. Thus, both the efficiency and the accuracy of focusing processing are guaranteed. Simulations are carried out to validate the proposed algorithm.

keywords: {Azimuth;Synthetic aperture radar;Focusing;Correlation;Spaceborne radar;Doppler effect;Frequency-domain analysis;Decoupled hybrid correlation (DHC);high-squint;synthetic aperture radar (SAR) imaging},

URL: https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10602522&isnumber=10354519

Paper Summary

Here is a summary of the key points from the paper:

- The paper proposes a new algorithm called Decoupled Hybrid Correlation Algorithm (DHCA) for processing high-squint spaceborne synthetic aperture radar (SAR) data in sliding spotlight mode.

- Key components of DHCA:
1) Sub-aperture processing to handle Doppler spectrum back-folding
2) Decoupled Hybrid Correlation (DHC) focusing method, including:
   - Range-Azimuth Decoupling (RAD) in 2D frequency domain
   - Nonlinear Chirp Scaling (NLCS) in range direction  
   - Consistent coarse range compression
   - Modified Hybrid Correlation (MHC)
3) Azimuth scaling

- Advantages of DHCA:
- Can process high-squint (45°) SAR data with 0.5m resolution in range and azimuth
- Maintains "azimuth-invariant" property, allowing batch processing
- More accurate focusing than existing algorithms, especially for targets away from scene center

- Simulation results show DHCA outperforms existing methods like Linear Range Walk Correction (LRWC) based algorithms in terms of focusing quality, especially for high-squint angles and targets away from scene center.

- DHCA has higher computational complexity than some existing algorithms, but is still relatively efficient.

- The paper validates DHCA through point target and distributed target simulations.

In summary, DHCA enables accurate and efficient processing of high-resolution, high-squint spaceborne SAR data over large scenes, overcoming limitations of existing algorithms.

Authors

Based on the information provided in the paper, here's what I can tell you about the authors, their associated institutions, and some related prior work:

Authors and Institution:

All six authors are affiliated with the School of Electronics and Information Engineering at Beihang University (also known as Beijing University of Aeronautics and Astronautics, BUAA) in Beijing, China. This suggests a strong research group in SAR technology at this institution.

1. Yanan Guo: Graduate Student Member of IEEE, pursuing a Ph.D. at Beihang University.

2. Pengbo Wang: Member of IEEE, an Associate Professor at Beihang University. He has prior experience as a visiting researcher at the University of Sheffield, UK.

3. Zhirong Men: Member of IEEE, an Assistant Professor at Beihang University. He also has experience as a visiting researcher at the University of Sheffield, UK.

4. Tao He: Graduate Student Member of IEEE, pursuing a Ph.D. at Beihang University.

5. Tian Qiu: Graduate Student Member of IEEE, pursuing a Ph.D. at Beihang University.

6. Jie Chen: Senior Member of IEEE, a Professor at Beihang University. He also has experience as a visiting researcher at the University of Sheffield, UK.

Prior Related Work:

While the paper doesn't extensively discuss the authors' prior work, it does reference several related algorithms and techniques that form the background for this research:

1. Linear Range Walk Correction (LRWC) method: Mentioned as a classic approach to handling high-squint SAR data processing.

2. Range Nonlinear Chirp Scaling (RNLCS) method: Another approach for processing high-squint SAR data, which this paper aims to improve upon.

3. Extended Nonlinear Chirp Scaling Algorithm (ENLCSA): A previous algorithm that used LRWC processing.

4. Extended Two-Step Focusing Approach (ETSFA): An earlier method for resolving 2-D spectrum distortion in high-squint SAR data.

The authors seem to have built upon these existing techniques to develop their new Decoupled Hybrid Correlation Algorithm (DHCA). The paper mentions that some of the authors have previously worked on topics such as high-resolution spaceborne SAR image formation, novel techniques for spaceborne SAR systems, multimodal remote sensing data fusion, and ionospheric effects on low-frequency space radars.

The repeated mentions of visiting research positions at the University of Sheffield suggest an ongoing collaboration or knowledge exchange between Beihang University and the University of Sheffield in the field of SAR technology.

Figures and Tables

Here's a list of the figures and tables in the paper, along with explanations of what each aims to show:

Figures:

1. Figure 1: Geometry model of the high-squint spaceborne SAR.
   - Illustrates the geometric configuration of the SAR system, defining key parameters and angles.

2. Figure 2: Block diagram of the proposed DHCA.
   - Provides an overview of the algorithm's structure and processing steps.

3. Figure 3: Echo signal in 2-D time domain for central range cells before and after LRWC processing.
   - Demonstrates the effect of LRWC on the echo signal.

4. Figure 4: Echo signal in 2-D time domain after RAD processing for central and non-central range cells.
   - Shows the impact of RAD processing on different range cells.

5. Figure 5: Range compression results for different range cells in the RD domain before NLCS.
   - Illustrates the need for NLCS by showing compression results before its application.

6. Figure 6: Range compression results for different range cells in the RD domain after NLCS.
   - Demonstrates the improvement in range compression after applying NLCS.

7. Figure 7: Construction block diagram of the reference function.
   - Outlines the process for constructing the reference function used in the algorithm.

8. Figure 8: Echo data extraction results along the residual RCM in the RD domain.
   - Shows the effectiveness of the data extraction method along the residual range cell migration.

9. Figure 9: Layout of the point targets utilized in the simulation.
   - Displays the arrangement of point targets used for algorithm validation.

10-14. Figures 10-14: Various simulation results and comparisons.
   - Present the outcomes of point target simulations, comparing DHCA with other algorithms.

15-18. Figures 15-18: Distributed target simulation results.
   - Show the performance of DHCA and comparative algorithms on distributed targets.

Tables:

1. Table I: Simulation Parameters
   - Lists the key parameters used in the simulations.

2. Table II: Parameter Values
   - Provides specific values for parameters used in computational complexity analysis.

3. Table III: Evaluation Results of the Proposed DHCA
   - Presents quantitative performance metrics for DHCA on different point targets.

4. Table IV: Evaluation Results of the LRWC-Based Algorithm
   - Shows performance metrics for the LRWC-based algorithm for comparison.

5. Table V: Evaluation Results of the RNLCS-Based Algorithm
   - Provides performance metrics for the RNLCS-based algorithm for further comparison.

These figures and tables collectively aim to illustrate the theory behind DHCA, demonstrate its implementation, and provide quantitative and qualitative comparisons with existing algorithms to prove its superiority in processing high-squint SAR data. 

DHCA Algorithm Functional Block Diagram

I'll explain each phase of the algorithm as shown in Figure 2, which illustrates the block diagram of the proposed Decoupled Hybrid Correlation Algorithm (DHCA). The algorithm consists of three main components:

1. Sub-aperture Processing:
   - Purpose: To handle the back-folding problem of the Doppler spectrum in the azimuth frequency domain.
   - Steps:
     a. Sub-aperture partition: Divides the full aperture into smaller sub-apertures.
     b. Nonlinear shift: Applies a frequency shift in the range frequency domain.
     c. Azimuth FFT: Performs Fast Fourier Transform in the azimuth direction for each sub-aperture.
     d. Time delay compensation: Compensates for the time delay caused by sub-aperture partitioning.
     e. Sub-aperture recombination: Combines the processed sub-apertures to form a dealiased 2-D spectrum.

2. Decoupled Hybrid Correlation (DHC):
   This is the main body of the algorithm, consisting of several steps:

   a. Range-Azimuth Decoupling (RAD):
      - Performed in 2-D frequency domain.
      - Eliminates the majority of range cell migration (RCM) and range-azimuth coupling (RAC) caused by high-squint angles.
      - Doesn't introduce extra range cell shifts, preserving the "azimuth-invariant" property.

   b. Nonlinear Chirp Scaling (NLCS) in Range Direction:
      - Equalizes the variant range chirp rate caused by residual RCM and RAC.
      - Prepares the signal for consistent coarse range compression.

   c. Coarse Range Compression:
      - Performs initial range compression uniformly for all targets.
      - Reduces the length of the subsequent Modified Hybrid Correlation (MHC) window.

   d. Modified Hybrid Correlation (MHC):
      - Extracts signals along the residual RCM trajectory.
      - Constructs a reference function in 2-D frequency domain.
      - Performs refined focusing, correcting residual RCM, compensating for residual Doppler phase modulation and RAC.
      - Achieves precise range compression for all range cells.

3. Azimuth Scaling:
   - Purpose: To eliminate the possible back-folded phenomenon in focused SAR images.
   - Steps:
     a. Applies an azimuth scaling function with a quadratic phase form.
     b. Performs azimuth FFT.
     c. Compensates for the consistent quadratic phase term.
     d. Performs azimuth IFFT to obtain the final focused image.

How the algorithm functions:

1. The input SAR data first undergoes sub-aperture processing to handle Doppler spectrum back-folding.
2. The DHC process then takes over:
   - RAD removes most of the RCM and RAC.
   - NLCS equalizes the remaining variant range chirp rate.
   - Coarse range compression is performed.
   - MHC refines the focusing process, correcting residual errors and achieving precise compression.
3. Finally, azimuth scaling is applied to produce a dealiased focused SAR image.

This algorithm allows for efficient and accurate processing of high-squint, high-resolution SAR data over large scenes by addressing the challenges of severe RCM and RAC while maintaining the ability to process data in batches in the azimuth frequency domain. 

Squint Angle Achieved

The DHCA algorithm was successfully demonstrated to process SAR data with a very high squint angle of 45°. This is a key achievement of the paper and represents a significant improvement over previous methods.

Specifically:

1. The point target simulations for the DHCA algorithm were conducted using a squint angle of 45°.

2. The algorithm achieved good focusing performance for all targets within the scene at this 45° squint angle, including targets at the edge of the imaging swath.

3. This 45° squint angle processing capability was maintained while achieving 0.5-meter resolution in both range and azimuth directions.

4. The paper states that under the simulation parameters used (listed in Table I), the DHCA can realize good focusing performance within an azimuth swath width of 10 km at this 45° squint angle.

5. For comparison, the LRWC-based and RNLCS-based algorithms used for comparison were only tested at a 15° squint angle, and showed degraded performance even at this lower angle.

The paper doesn't explicitly state an upper limit for the squint angle that can be processed using DHCA. However, the 45° angle demonstrated is already considered a very high squint angle for spaceborne SAR systems, representing a significant advance in processing capabilities for highly squinted SAR data.

This high squint angle capability allows for more flexible earth observation, faster revisit times, and the ability to image areas that might be challenging to observe with more traditional, less-squinted SAR geometries.