### Recommended for you

Online Available online. Full view. SAL3 off-campus storage. S95 C37 Available. More options. Find it at other libraries via WorldCat Limited preview. Contributor Goodman, Ron S. Majewski, Ronald M. Bibliography Includes bibliographical references and index. Part 3 Spotlight SAR and polar format algorithm: scope of processing task-- polar format overview-- polar data storage as a two-dimensional signal-- correction for non-planar motion-- polar format algorithm limitations-- Taylor series expansion procedures-- phase of image pixels-- image geometric distortion-- image focus error equations-- displacements and absolute positioning.

Part 4 Digital polar format processing: sampling rate conversion-- polyphase filters-- polar interpolation-- image scale factors-- image distortion correction-- signal history projections-- stabilized scene polar interpolation-- subpatch processing and mosaicking. Part 5 Phase errors: classification of phase error-- management of phase error-- magnitude of phase error-- requirements on a practical SAR motion sensor-- moving target effects. Part 6 Autofocus techniques: mapdrift-- multiple aperture mapdrift-- phase difference-- phase gradient-- prominent point processing-- considerations for space-variant refocus.

Part 8 SAR system performance: image quality metrics-- system performance budgeting-- requirements on system impulse response-- requirements on system noise-- geometric distortion-- secondary image quality metrics-- test arrays. Part 10 Range migration algorithm: model-- algorithm overview-- analytical development-- discussion-- efficient algorithms for range migration processing. Part 11 Chirp scaling algorithm: non-dechirped signal model-- algorithm overview-- analytical development-- discussion.

By author s: Walter C. Carrara , Ron S. Goodman , Ronald M. Description Contents Author Reviews Authored by engineers for engineers, this book is designed to be a practical and easy-to-understand solution sourcebook for real-world high-resolution and spot-light SAR image processing. Widely-used algorithms are presented for both system errors and propagation phenomena as well as numerous formerly-classified image examples. As well as providing the details of digital processor implementation, the text presents the polar format algorithm and two modern algorithms for spot-light image formation processing - the range migration algorithm and the chirp scaling algorithm.

Bearing practical needs in mind, the authors have included an entire chapter devoted to SAR system performance including image quality metrics and image quality assessment. Another chapter contains image formation processor design examples for two operational fine-resolution SAR systems. This is a reference for radar engineers, managers, system developers, and for students in high-resolution microwave imaging courses.

The proposed imaging process is accurate and highly efficient for sliding spotlight SAR mode. Stripmap mode can achieve wide swath imaging with the azimuth resolution below a half of azimuth antenna length [ 1 ]. Spotlight mode is characterized by azimuth antenna steering to a rotation center, which can achieve high azimuth resolution and a narrow swath. Sliding spotlight mode is characterized by azimuth antenna steering to a virtual rotation center during the raw data acquisition interval. With the control of the beam rotation rate and raw data acquisition interval, high resolutions and wide swath observation can be achieved simultaneously.

In other words, sliding spotlight mode is the trade-off between stripmap mode and spotlight mode. The sliding spotlight mode in TerraSAR-X [ 11 ] applies a slower antenna steering than staring spotlight mode. The launch of the Chinese satellite Gaofen-3 in is the first C-band and high-resolution, fully polarimetric SAR satellite with 12 imaging modes, including sliding spotlight mode [ 12 , 13 , 14 ].

Different from traditional stripmap SAR, signal processing of sliding spotlight echoes faces several difficulties. Second, for the resolution is high in sliding spotlight imaging, and the Doppler parameter must be acquired accurately either via orbit parameters in the WGS coordinates system or Doppler parameter estimation.

### See a Problem?

Various algorithms used to overcome spectrum aliasing have been proposed in recent years [ 15 , 16 , 17 , 18 , 19 , 20 ]. The classical two-step imaging algorithm is introduced to overcome the azimuth spectrum aliasing [ 15 , 16 ]. Sub-aperture algorithms are another way to deal with the azimuth spectrum aliasing [ 17 ], whereas the sub-aperture algorithms need sub-aperture formation, which is not efficient due to azimuth data overlap. These methods can overcome spectrum aliasing and achieve sliding spotlight imaging. However, the Doppler parameter estimation methods are required to combine with the imaging algorithm due to the high-resolution imaging.

In principle, it is possible to calculate the Doppler parameter from orbit and attitude data, but measurement uncertainties on these parameters will limit the accuracy. In order to obtain the accurate Doppler parameter, a lot of Doppler parameter estimation methods are proposed in recent years.

The Doppler centroid can be estimated by [ 21 , 22 , 23 , 24 ]. References [ 21 , 22 ] propose Doppler centroid estimation algorithm using azimuth spectrum and antenna pattern, which are kinds of frequency algorithm. The proposed algorithm in references [ 23 , 24 ] is a time domain algorithm and denoted the correlation Doppler estimator with high efficiency. However, these methods are suitable for Doppler centroid estimation in stripmap mode since the PRF is slightly above the Doppler bandwidth.

Azimuth antenna steering in sliding spotlight leads to the variation of the Doppler centroid in azimuth, and the Doppler bandwidth exceeds PRF, which cannot be applied in traditional Doppler centroid method.

## Spotlight Synthetic Aperture Radar: Signal Processing Algorithms

The Doppler frequency rate is one of the key parameters in the azimuth focusing of SAR data. Once mismatched, it will cause serious defocusing in azimuth direction and result in the degradations of image quality. Several Doppler frequency rate estimation algorithms [ 25 , 26 , 27 , 28 , 29 , 30 ] are proposed including the Map Drift MD [ 25 , 26 ], phase gradient autofocus PGA [ 27 , 28 ] and contrast optimization algorithm [ 29 ] with high accuracy.

However, since different azimuth targets share different time support domains and different frequency support domains in sliding-spotlight mode, the MD algorithm cannot be applied as the Doppler frequency rate estimation algorithm. Also, the PGA algorithm cannot be applied in the sliding spotlight working mode directly since different azimuth targets share different time support domain, which is different from spotlight mode. In this paper, contrast optimization algorithm is introduced and adopted as Doppler frequency rate estimation.

In this paper, we proposed an integrated imaging scheme with sliding spotlight mode. In the imaging scheme, the two-step approach is firstly applied to the spaceborne sliding spotlight SAR imaging algorithm, followed by the modified correlation Doppler centroid estimator and modified contrast optimization algorithm.

The azimuth spectral fol algorithm ding phenomenon is overcome by the two-step approach, and the chirp scaling CS [ 1 ] algorithm is applied to obtain the unambiguous images.

The Doppler centroid variation along azimuth can be estimated by modified correlation Doppler estimator with high accuracy. The Doppler frequency rate variation along range can be estimated by contrast optimization algorithm with high accuracy. The proposed imaging process is accurate and efficient for sliding spotlight SAR mode. This imaging mode provides 1 m resolution with 10 km swath width.

The planar imaging geometry of the sliding-spotlight mode is shown in Figure 1 , which is simplified by linear geometry since the azimuth rotation angle is small enough. In the sliding spotlight imaging geometry, the azimuth beam steers from fore to aft at a constant rotation rate as.

V f is the footprint velocity taking into account azimuth beam steering. Based on rectilinear imaging geometry of sliding-spotlight mode, R t a can be written as. Because of the Range Doppler RD [ 1 ], CS and nonlinear CS NCS [ 31 , 32 ] algorithms are all high efficient frequency imaging algorithm without interpolation, and one of the imaging algorithm is adopted in the Gaofen-3 sliding spotlight processing. The major error in CS algorithm CSA including two parts: one is the variation of V r along the range direction, which leads to RCM error, and another is the variation of K m in range-Doppler domain, which is assumed not related to range [ 1 ], where K m can be expressed as.

Firstly, the effective velocity V r slightly varies with range. Thus, the RCM error caused by V r is independent of the range direction and can be expressed as. The RCM error is caused by the effective velocity V r slightly varying with range. Secondly, the K m error caused by it varies with R 0 is calculated by. In essential, this error is the variation of second range compression SRC in range time domain which cannot be compensated in CS algorithm. The variation between SRC error and range frequency is presented in Figure 3.

Notice f a in Equation 8 is chosen as maximum of Doppler bandwidth.

## Spotlight synthetic aperture radar signal processing algorithms

The SRC error is caused by R 0 variation in time domain. In this section, based on CSA, we propose a complete process of sliding spotlight imaging with Doppler parameter estimation. Figure 4 is the processing flow of the Gaofen-3 sliding spotlight mode. Processing overview for Gaofen-3 sliding spotlight imaging.

The main contributions of this paper are the Doppler centroid estimation and Doppler frequency rate estimation, which are shown in yellow box.

From Figure 4 , three techniques are vital for obtaining focused images: Deramp operation, Doppler centroid estimation, and Doppler frequency rate estimation. As Figure 4 shows, the Doppler centroid estimation and Doppler frequency rate estimation are the main contributions of this paper. In general, the total Doppler bandwidth is larger than the PRF in the sliding spotlight mode. In order to overcome the aliasing of the azimuth echo in the frequency domain, azimuth convolution processing is conducted, which is the key point of azimuth preprocessing.

The quadratic phase signal is expressed as. While conducting the azimuth weighting processing, the azimuth preprocessing can be accomplished by employing. From Equation 11 , the azimuth convolution processing includes three parts: dechirp processing, Fourier transform, and residue compensation.

A lot of Doppler centroid estimation algorithms are developed in SAR imaging. The proposed algorithm in reference [ 23 ] is denoted the correlation Doppler estimator CDE with high efficient. In Ref. Notice in Equation 13 , the summation should not be along the azimuth direction since the azimuth beam steering leads to the variation of f d c , which should be estimated.

To improve the accuracy in low SNR condition, taking the average in range is necessary and helpful. Therefore, the variation of f d c in range direction is neglected, which leads to a small error about several hertz. The variation of f d c in range direction will be discussed below. Because the f d c in azimuth beam steering may be greater than PRF, Doppler ambiguity occurs.

Fortunately, the sliding spotlight mode in GF-3 is not squinted SAR, and thus the Doppler ambiguity number in azimuth reference time is 0.