Title of Invention	A METHOD FOR PROCESSING SPACEBORNE SLIDING SPOTLIGHT SYNTHETIC APERTURE RADAR SIGNAL FOR EXTENDED AZIMUTH COVERAGE
Abstract	The invention provides a method of processing for extending azimuth coverage for a SAR operating in spaceborne sliding spotlight mode. Two major difficulties, in processing such data, namely large range cell migration (RCM) and rapidly changing Doppler rate along azimuth. This method assumes that the complete azimuth stretch will be processed in different blocks. Majority of RCM will be corrected in time domain by Bulk RCM correction and a corresponding phase compensation will convert the data to a strip-map one. However, in the resultant stripmap data, the Doppler rate drifts along the azimuth and the drift is prominent as one moves away from spot center. A resampling mechanism, Co applicable in both time and frequency domain, depending upon the matched filter domain, ,is suggested which enables stripmap processing over a block of data with linearly drifting Doppler rate.

Title of Invention

A METHOD FOR PROCESSING SPACEBORNE SLIDING SPOTLIGHT SYNTHETIC APERTURE RADAR SIGNAL FOR EXTENDED AZIMUTH COVERAGE

Abstract

The invention provides a method of processing for extending azimuth coverage for a SAR operating in spaceborne sliding spotlight mode. Two major difficulties, in processing such data, namely large range cell migration (RCM) and rapidly changing Doppler rate along azimuth. This method assumes that the complete azimuth stretch will be processed in different blocks. Majority of RCM will be corrected in time domain by Bulk RCM correction and a corresponding phase compensation will convert the data to a strip-map one. However, in the resultant stripmap data, the Doppler rate drifts along the azimuth and the drift is prominent as one moves away from spot center. A resampling mechanism, Co applicable in both time and frequency domain, depending upon the matched filter domain, ,is suggested which enables stripmap processing over a block of data with linearly drifting Doppler rate.

Full Text	FIELD OF INVENTION : This invention relates to a method for efficiently processing sliding spotlight synthetic aperture radar (SAR) signal. The efficiency of processing is increased by maximizing the block size for processing. The invention provides a method of processing sliding spotlight SAR signal with extended azimuth coverage for a spacebome SAR operating in sliding spotlight mode. There are two major difficulties in processing such data, due to large range cell migration (RCM) and rapidly changing Doppler rate along azimuth. The method assumes that the complete azimuth stretch will be processed in different blocks. Each of the block will be stretched both along azimuth and range. This method will allow maximizing the blocksize along azimuth. Majority of RCM will be corrected in time domain by bulk RCM correction and a corresponding phase compensation will convert the data to a stripmap one. However, the bulk RCM correction will not correspond to any physical target on the scene, but will correspond to the trace of assumed beam centre line. In the resultant stripmap data, the Doppler rate drifts along the azimuth and the drift is prominent as one moves away from spot center. A resampling mechanism, applicable in both time and frequency domain, depending upon the matched filter domain, is suggested which enables stripmap processing over a block of data with linearly drifting Doppler rate. This method substantially enhances the azimuth blocksize in comparison to existing processing methods for processing stripmap SAR data. The compressed sliding spotlight SAR data in azimuth time slant range domain is provided a reverse bulk RCM correction. Now in the compressed data, each azimuth line corresponds to one distinct beam center Doppler frequency. From the knowledge of this beam center Doppler frequency and slant range, each of the target can be appropriately mapped to ground coordinate system. PRIOR ART : Synthetic aperture radar (SAR), conceived by Carl A Willey in 1952, was originally conceptualized in stripmap imaging configuration. The best possible azimuth resolution in this imaging configuration is 1/2 where / is the antenna length along flight direction which is conventionally called as azimuth length. The limitation in the best possible resolution arises out of maximum limit of the length of synthetic antenna and is limited to physical size of the azimuth footprint of the beam. As far as imaging length in flight direction is considered, there is no physical limit imposed by SAR concept. Almost three decades later another concept of imaging in SAR called spotlight mode emerged. The maximum limit on the length of synthetic aperture is removed by steering the beam towards a particular spot being imaged. Theoretical best possible resolution achievable by spotlight mode is X/4 at infinite observation duration. With practical limits on observation duration, typical resolution obtained by spotlight imaging is -^lOX. On the flipside, the azimuth length of imaging is restricted within the azimuth footprint of the antenna beam. Primary differences of the above mentioned two imaging modes are summarized in Table-1. Table-1 Salient Differences in Stripmap and Spotlight Modes There are certain differences in the SAR processing strategies implemented in these two imaging modes. Two major processing steps range cell migration (RCM) correction and aperture synthesis are generally implemented in different ways. In stripmap case, as the net RCM is small enough for implementation of RCM correction in frequency domain, this correction is usually implemented in Doppler frequency space. Also as the azimuth data length to be processed is much larger than the matched filter length, matched filtering is usually done in frequency domain and image is obtained by inverse transformation in time domain. In spotlight mode, the RCM encountered is much higher making its correction in frequency domain difficult without phase distortion and the net output image is much smaller than the matched filter length. Consequently RCM correction is employed in time domain, corresponding to center target, usually on receiving. The matched filter is implemented by time domain phase multiplication whereby targets along azimuth direction gets separated in terms of distinct frequency tones, much akin to frequency modulated continuous wave FMCW radar processing. This process is usually termed in SAR literature as deramping. Image is obtained in frequency domain through one forward Fourier transform operation. This peculiar sort of RCM correction leads to best focusing at the center of the image and gradual degradation in focusing as one moves away from image center. They are subsequently sharpened by a secondary range compression operation. Salient differences in the processing strategies of these two modes are presented in Table-2. All the above discussions regarding comparison of these two modes corresponds to one SAR antenna being used for both strip map and higher resolution spot light operation. During 1994-1996, a new concept of SAR imaging appeared which promised to remove limitations of spotlight mode in achieving larger extent of azimuth stretch of imaging beyond azimuth footprint of the antenna. It is essentially a spotlight SAR with reduced ground speed of the beam trace of the antenna, whereby the aperture time is increased, sufficient enough to achieve improved resolution. This concept is described by D P Belcher and CJ Baker on pages 366-374 of lEE Proc.Radar Sonar Navigation Vol.143, No.6, December 1996. The processing strategy to be employed for processing of sliding spotlight mode of SAR is described by Riccardo Lanari, Manilo Tesauro, Eugenio Sansoti, Gianfranco Fomaro on pages 1993-2003 of IEEE Trans. On Geosci. and Remote Sensing,Vol. 39, No. 9, Sept. 2001 and on pages 363-372 lEE Proc.-Radar, Sonar Navigation Vol. 148, No. 6, Dec 2001. The processing of sliding spotlight data has been attempted with following strategy in these disclosures. 1. In order to correct high RCM two step RCM correction is employed. The first step is bulk RCM correction, identical to that employed in classical spotlight SAR whereby the RCM of the center target is compensated in time domain along with a phase compensation operation. This method converts the sliding spotlight data to stripmap data. Residual RCM is corrected in the frequency domain during stripmap processing. 2. Matched filtering is carried out in frequency domain similar to stripmap processing. In essence, existing literature has addressed the problem as an extension of classical spotlight processing and the processing exhibits the typical signature of spodight processing with best focusing at the center of the image and loss of focusing as one moves towards image periphery. This loss of focus is also compensated by a secondary compression. This treatment concludes stating that the present analysis is focused on spacebome systems typically characterized by small squint angles during acquisition (often less than 1°) and negligible motion errors. The extension of the proposed method to the airbome case is certainly possible but, in such case, additional and specific issues such as motion compensation and/or high squint angle values in the data acquisition must be carefully considered and may be worth pursing. As per the above conclusion, from space-bome SARs, operating at typical altitudes of 600 to 800 km it will be difficult to achieve sliding spotlight images larger than the area of the order of 10 kmxlOkm with resolution of the order of 1 m. If one attempts to extend the azimuth extend to 100 km with spacebome sliding spotlight SAR, azimuth squinting in excess of 10° will be required to achieve resolutions of the order of 1 m. As per the above conclusion, present available methods will not be able to process such data. The difficulties of processing sliding spotlight SAR data for extended azimuth coverage can be best understood from a specific example. Consider a hypothetical spacebome SAR system specified in Table-3 below. The major handicap in processing this data from this hypothetical spacebome SAR is widely varying Doppler parameters and RCM. From the typical variations in Doppler parameters and RCM, it is evident that histories of individual targets are limited within a Doppler band traced by fore and aft end of the beam. The beam center Doppler will vary between 0 Hz at center of the spot to around ±40000 Hz at azimuth extremities. However, the instantaneous Doppler bandwidth is within 2500 Hz. This fact enables PRF choice determined by the instantaneous Doppler bandwidth. The RCM behaviour shows widely varying linear RCM component to the tune of 4.5 km at the edges of azimuth extent. In real terms, it implies that phase history of one target can get spread over an area extending around 11000 samples along azimuth and 7500 samples along range. It should be noted that one can process SAR images by time domain correlation by defining matched filter for every individual pixel. But such a brute force processing strategy will be prohibitively expensive. Best computation performance is achieved when block processing is implemented wherein same matched filter is used for compressing all the targets along azimuth within a block. It is needless to mention that matched filter needs to be updated along range as azimuth match filter is a function of range. The block-size is to be chosen such that Doppler parameters are stable within a block. The stability criteria for Doppler rate is --0.003% for achieving optimum focusing and '-0.5% for optimum RCM correction. In all the processor design, block-size is chosen for meeting the Doppler rate criteria optimum for focusing. Such a choice will yield hardly any block-size as one moves away from spot center. At spot center, this criteria is easily met for considerable block-size. However, as one moves away from spot center, the azimuth size of the block gets smaller. The upper bound of the azimuth block size is determined by the stability criteria of Doppler parameters governed by RCM correction. Such a criterion will increase the block size in azimuth direction significantly. However, the processing method should be capable of processing the entire azimuth block by a single matched filter fiinction. OBJECT OF THE INVENTION : The object of invention is to develop a method for processing an azimuth block of data for sliding spotlight mode of a spacebome SAR which exhibits large drift in Doppler rate parameter within a block with a single matched filter fiinction. This enables achieving maximum block size from the relaxed requirement of stability of Doppler rate parameter for carrying out RCM correction. Consequently the method improves the net computation efficiency of processing. DESCRIPTION OF INVENTION: The method for processing spacebome sliding spotiight SAR signal for extended azimuth coverage is carried out in the following steps. Dividing the data in suitable blocks spread along azimuth and range assuming same Doppler centroid is assumed in the range extent of the block. Bulk RCM is corrected by shifting the target at range R to R-AR in accordance with the formula AR= -A/DC?/2- ^/DR_BEAM_CENTRE^^/^ wherein, /DC andyi)R BEAM CENTRE^-^ ^^^ Doppler centroid and beam center Doppler rate. The same bulk RCM correction factor is used for all the range gates within a block of data. After bulk RCM correction, a phase compensation is carried by multiplying the bulk RCM corrected signal with the following function : exp(-i27i/Dct -infm BEAM CENTRE^^) The same phase compensation factor is used for all the range gates within a block of data. Typically the bulk RCM correction and phase compensation coefficients will be chosen for the target situated at the center of the block. By carrying out steps 2 and 3, the data is converted with a phase history equivalent to a stripmap SAR with a difference. In the stripmap SAR all the phase histories of all the targets lying in the same range gate exhibit same Doppler rate. However, in this case the phase histories of all the. targets lying in the same range gate after bulk RCM correction will exhibit different Doppler rate. Further the target from the target situated at center of azimuth block, the drift in Doppler rate will be more. The net drift in Doppler rate will be least for the block situated at the spot center at which antenna is looking in broadside direction. The drift in Doppler rate will be more if the block is situated away from spot center when antenna is looking at the target with more azimuth squint angle. A method of block processing of resultant data with linearly drifting Doppler rate along azimuth is carried out in accordance with the method involves resampling of raw data, either in time domain for deramping based processor or in frequency domain for matched filtering in frequency domain. This resampling scheme converts the target histories along azimuth from drifting Doppler rate to constant and identical phase history. This resampling scheme enables the processing of resultant stripmap data with single matched filter frmction. Resampling scheme in time domain for deramping based processor is carried out as follows. i. Residual RCM is carried out in frequency domain by taking short burst of data in frequency domain and then bringing back the data to time domain after residual RCM correction. ii. The raw data is multiplied with matched filter history corresponding to center target of azimuth block. iii. The resultant phase histories are alligned by convolving with a filter to obtain an image in frequency domain. iv. Resultant data is resampled to convert phase histories of individual targets from a slow chirp to constant frequency signal. V. Final image is obtained i frequency domain. Resampling scheme in frequency domain for frequency domain matched filtering is carried out as follows. i. The raw data is converted to frequency domain. ii. Residual RCM is corrected in frequency domain. iii. The azimuth spectrum is resampled as per resampling scheme to convert varying phase histories of different targets to identical phase history for all targets, iv. The time domain history of center target in azimuth block is converted to frequency domain and identical resampling procedure is adopted. V. Resampled raw data in frequency domain is multiplied with complex conjugate of resampled phase history of the center target, vi. Inversion of the matched filtered signal from frequency to time domain to obtain image. 5. The compressed image data is subjected to reverse bulk RCM correction by shifting target at range R to R+AR wherein AR- -^fY)Qt/2- /iyi)R BEAM CENTRE^^/"^ In the image domain, image points at a particular azimuth time, for all range gates, correspond to one particular beam center Doppler frequency and hence at one azimuth squint angle. From this information, all the targets at particular azimuth time can be repositioned at ground coordinate system. DESCRIPTION OF FIGURES : Fig.-l shows a pictorial description of imaging operation of conventional strpmap SAR. Fig.-2 shows a pictorial description of imaging operation of conventional spotUght SAR. Fig.-3 shows a pictorial description of imaging operation of sliding spotlight SAR. Fig.-4 shows a typical behaviour of Doppler graphic representation of the parameters for the hypothetical spacebome sliding spotlight SAR over 100 km azimuth extent of imaging. Fig.-5 shows a typical behaviour of range cell migration parameters for the hypothetical spacebome sliding spotlight SAR over 100 km azimuth extent of imaging. Fig.-6 shows a pictorial depiction of translation of Doppler phase history because of bulk RCM correction and phase compensation. Fig.-7 illustrates range history of targets which appear at the same range gate after bulk RCM correction. Fig.-8 shows a pictorial depiction of behaviour of phase histories of the targets with drifting Doppler rate during various stages of modified deramping processor scheme incorporating resampling in time domain. Fig.-9 shows resampling scheme in time domain for deramping base SAR processing. Fig.-10 shows block schematic of modified reramping based SAR processor incorporating time domain interpolation for processing sliding spot light signal. Fig.-l 1 shows improvement in impulse response behaviour of the targets spread over 1 second of azimuth flight time at one end of 100 km azimuth extent with conventional and modified deramping processor scheme incorporating time domain resampling. Fig.-12 shows a pictorial depiction of behaviour of phase histories of the targets with linearly drifting Doppler rate during various stages of modified frequency domain matched filtering processor scheme incorporating resampling in frequency domain. Fig.-13 illustrates implementation scheme for implementation in frequency domain matched filter convolution approach to account for linear drift in azimuth chirp rate. Fig.-14 shows a block diagram of modified matched filter convolution based SAR processor for processing stripmap SAR data with drifting Doppler rate. Fig.-15 shows an improvement in impulse response behaviour of the targets spread over 1 second of azimuth flight time at one end of 100 km azimuth extent with conventional and modified frequency domain matched filtering processor scheme incorporating frequency domain resampling. Fig.-16 shows a two dimensional simulation of modified matched filter convolution method with RISAT parameters at 50 km away from spot centre, showing improvement in impulse response (a) with conventional method (b) with modified method. Fig.-17 shows geometric reference of reverse bulk RCM corrected processed image. Fig.-18a to 18g shows a two dimensional simulation result for an array of point targets placed at one end of 100 km azimuth extent. Fig.-19 shows impulse responses of the three targets, shown in Fig. 18(g), after mapping to ground coordinates, the Z-axis is being in dB scale. Fig.-20 shows relationship of processing filter bandwidth location vis-avis doppler phase history for a set of targets. (a) Without any error in estimation of azimuth steering rate and doppler centroid of centre target. (b) With error in estimation of azimuth steering rate and no error in estimation of doppler centroid of centre target. (c) With a error in estimation of both azimuth steering rate and doppler centroid of centre target. Fig.-21 shows a demonstration of resilience of the processor developed by the method according to the invention with accurate and erroneous the knowledge of scan rate. Thus the method according to the invention incorporates processing stripmap SAR data with linearly drifting Doppler rate along azimuth via a resampling process either in time domain (for deramping based processor) or in a frequency domain (for frequency domain matched filtering based processor). The effectiveness of this method is demonstrated by two processing approaches in time as well as frequency domain. The novel resampling scheme ensures that the phase histories of all the targets along azimuth, lying at a particular range gate after bulk RCM correction, will become identical for all these targets. Thus block processing by a single matched filtering is possible. The increase in block size can be seen in the simulation results presented in Figs. 11 and 15. Possibility of increased block size ensures higher computation efficiency. The azimuth size of the block is govemed by the lower size dictated by the linearity in the drift rate of the Doppler rate or by the net drift in Doppler rate which can be accommodated in residual RCM correction. The range size of the block is govemed by the rate of change of Doppler centroid along range as same Doppler centroid figure is assumed for one block. Though Bulk RCM correction is a conventional process, in all the existing methods, the bulk RCM correction corresponds to the block center target which has a physical existence. However, in the present method, the bulk RCM correction factor does not correspond to any physical target. It corresponds to trajectory of the beam center line which generates linear Doppler history. This is a novel concept. Conventional stripmap processing or spotlight processing can be arrived at in the suggested method by very minor modifications in the coding. Bulk RCM correction and phase compensation scheme used in the method according to the invention is described below. compensation and bulk RCM correction will drift in proportion to the time difference with respect to the block center target. The drift rate in Doppler rate is increased as one moves away from spot center along azimuth. Subsequent discussion is concentrated on block processing the data where azimuth Doppler rate of the target is drifting linearly. Further, higher order phase component will also change. The block size will get limited by the condition that the drift in Doppler rate is linear and the change in higher order phase term over the block length is not appreciable. Stripmap processing with drifting Doppler rate scheme used in the method according to the invention is described below: Block processing of stripmap SAR is possible if Doppler rate is constant or the phase variation due to Doppler rate drift is within tolerable limits of n/4. Beyond this limit the impulse response degrades and this constraint is the only limiting factor in determining the block size. In the following sections an approach is presented for block processing of strip map SAR with drifting Doppler rate. This method will extend the azimuth block size significantly limited by the constraint that linearity of drift rate holds good over the azimuth block extent. Two different implementations are provided. One using deramping method and the other using matched filtering method, are presented for demonstrating the effectiveness of this approach. ^/J^] = ^/J^-l] + A7}4v[^-l])--(25) The resampling is accomplished by equating output and input signals as Fig." 10 shows the complete processing scheme of sHding spotHght SAR signal using deramping based processor. Fig." 11 shows effectiveness of this method with one dimensional simulation. For this simulation hypothetical spacebome SAR parameters shown in Table-3 at 50 km away (azimuth) from the spot center and 560 km ground distance from nadir track were assumed. The foj^ MODIFIED =' 715 Hz/sec and /OR RATE ='^'^ Hz/sec were assumed. Conventional stripmap processing would not have produced sufficient block image there by drastically reducing block processing efficiency. Because of the above mentioned modification, even at the far end, a block data may be processed over 1 second flight duration which is a significant achievement. The duration over which the processor can be applied depends upon/^^ ^^^^ and residual cubic phase term. In a frequency domain matched filter convolution processing of stripmap SAR data with drifting Doppler rate like the previous example of deramping processing by resampling in time domain, the corresponding processing using frequency domain matched filter convolution is also possible by resampling the signal in frequency domain. The processing methodology is presented in Fig.-12. Let us consider three targets as shown in Fig.-12(a), one at the center of the aperture with beam center at / = 0 and two aft and forward targets with Fig.-15 shows effectiveness of this method with one dimensional simulation. For this simulation hypothetical spacebome SAR parameters (Table-3) at 50 km away (azimuth) from from spot center and 560 km ground distance from nadir track were assumed. The /^.^ MODIFIED =-715 Hz/sec and /^^^ j^^^ =-6.6 Hz/sec were assumed. Conventional stripmap processing would not have produced sufficient block image there by drastically reducing block processing efficiency. Because of the above mentioned modification, even at the far end we can process a block data over 1 second flight duration which is a significant achievement. The duration over which the processor can be applied depends upon/^;; ^^^j^ and residual cubic phase term. Fig.-A.16 shows the effectiveness of this method with two dimensional point target simulation. Geometric distortion correction scheme used in the method according to the invention is described below. The first step to geometric correction after azimuth compression is reverse bulk RCM correction. After this reverse bulk RCM correction, the image is mapped from Azimuth time - slant range domain to ground coordinate system (x,y) with the origin of reference being kept at nadir location at / = 0 reference. Fig.-17 is drawn with flat earth approximation for ease of illustration of salient points in Geometric correction. The matched filter is generated at t = t^ reference with Doppler Centroid IDC (^0) • It is to be noted that in the processed image in Azimuth time- Computer simulation scheme of an array of point targets used in the method according to the invention is given below. The sliding Spotlight processor has been implemented with matched filter convolution. Raw data has been simulated for the hypothetical SAR configuration shown in Table-3 for a set of point targets with a matrix arrangement of 50mx50m spacing spread over 500mx500m area at a ground range distance of 560 km from nadir track. The targets are placed at 50 km distance, opposite to velocity vector direction, from the spot center of spot size 10km (Range)xlOO km (Azimuth). During raw data simulations, no approximation has been made in computing slant range and hence phase history of the targets as the satellite passes over. From raw data to output ground co-ordinate image stage, all the transformations in the data are shown in Fig.-18. Impulse responses of the marked targets in Fig.-18(g) is shown in Fig.-19. One interesting observation needs to be pointed out that the method is resilient to Doppler centroid estimation error and estimation error in pitch rate. In fact the current depiction of simulated data has been processed with Doppler centroid estimation error of 200 Hz and 10% error in the estimation of pitch rate. The reason of resilience of the processor is illustrated in Fig."20. In this figure the signal Doppler History and Processor Bandwidth location is presented for a set of three targets. There is an exaggeration in this figure: in reality, the Doppler histories are more closer and almost overlapped and processor bandwidth is much larger and almost encompassing the signal bandwidth. However, this exaggeration makes the illustration simpler. This explains even in presence of estimation in Doppler centroid and steering error, image will still be processed if the processing filter is located in Doppler phase history. For different errors, the targets will appear at different azimuth time and different slant ranges. However, after geometric correction, the targets will appear at the correct ground coordinates. This effect is illustrated in Fig.-21. This figure shows how the targets are positioned at different azimuth time and slant range after processing and reverse bulk RCM correction for different error in estimation of pitch scanning rates. However, in both cases the targets are correctly located in ground coordinate system after geometric correction. The method according to the invention, is intended to increase the azimuth extent for sliding spotlight SAR imaging from a space-bome platform. Usually all practical SAR processors are designed with processing over a number of blocks of raw data and consequent seamless stitching of individually processed blocks. As one moves from image center in sliding spotlight imaging, the Doppler parameters and RCM change progressively faster over a block of data. The conventional SAR processing, which requires stability of Doppler parameters over a block, will not lead to appreciable block size as one moves away from image center, leading to inefficient processor architecture. The present method addresses the issue of maximizing the block size even under this condition. The heavy RCM is maximally corrected in this method by bulk RCM correction in time domain. However, this bulk RCM correction does not correspond to a physical target in the image, but to the beam center trace. A corresponding phase compensation converts the data to a stripmap one and the residual RCM is corrected in the stripmap processing. However, the resultant stripmap raw data, so obtained, exhibits the same behaviour of progressively higher drift in Doppler parameters, as one moves away from image center. A modification in matched filtering method is suggested to take care of such drifts within a block as long as the drifts in Doppler rate is linear. The suggested method is one method for interpolation of the resultant stripmap raw data in time domain or frequency domain, depending upon implementation domain of matched filtering. Further, geometric mapping of the final image from azimuth time - slant range reference frame to ground coordinate reference frame has been investigated. The resilience of the method, even in the presence of Doppler parameter estimation error, to map the image accurately in ground coordinate system has been demonstrated. Apart from analytical development, the method has been demonstrated by processing simulated raw data of a matrix of point targets. For simulation of the raw data, hypothetical SAR parameters (Table-3) were assumed. WE CLAIM : A method for processing spacebome sUding spot hght synthetic aperture radar (SAR) signal for extended azimuth coverage comprising the steps of: (i) dividing the data in pre-determined blocks spread along azimuth and range assuming same Doppler centroid in the range extent of the block; (ii) converting bulk range cell migration (RCM) in accordance with the bulk RCM correction scheme described herein using same bulk RCM correction factor for all the range gates within a block of data; (iii) compensating the phase by multiplying the bulk RCM corrected signal in accordance with the phase compensation scheme described herein using the same phase compensation factor for all range gates within a bock data; (iv) stripmap processing with drifting Doppler rate in accordance with the stripmap processing scheme described herein; (v) correcting geometric distortion in accordance with the geometric distortion correction scheme described herein; and (vi) obtaining an array of point targets using computer simulation scheme of an array of point targets described herein. The method as claimed in claim 1, wherein the said stripmap processing step is carried out by deramping based processing of stripmap SAR data with drifting Doppler rate as herein described. The method as claimed in claim 1, whereien the said stripmap processing step is carried out by frequency domain matched filter convolution processing of stripmap SAR data with drifting Doppler rate as described herein. A method for processing spacebome sliding spot light synthetic aperture radar (SAR) signal for extended azimuth coverage substantially as hereinabove described and illustrated with reference to the accompanying drawings.

Full Text

FIELD OF INVENTION :
This invention relates to a method for efficiently processing sliding spotlight synthetic aperture radar (SAR) signal. The efficiency of processing is increased by maximizing the block size for processing.
The invention provides a method of processing sliding spotlight SAR signal with extended azimuth coverage for a spacebome SAR operating in sliding spotlight mode. There are two major difficulties in processing such data, due to large range cell migration (RCM) and rapidly changing Doppler rate along azimuth. The method assumes that the complete azimuth stretch will be processed in different blocks. Each of the block will be stretched both along azimuth and range. This method will allow maximizing the blocksize along azimuth. Majority of RCM will be corrected in time domain by bulk RCM correction and a corresponding phase compensation will convert the data to a stripmap one. However, the bulk RCM correction will not correspond to any physical target on the scene, but will correspond to the trace of assumed beam centre line. In the resultant stripmap data, the Doppler rate drifts along the azimuth and the drift is prominent as one moves away from spot center. A resampling mechanism, applicable in both time and frequency domain, depending upon the matched filter domain, is suggested which enables stripmap processing over a block of data with linearly drifting Doppler rate. This method substantially enhances the azimuth blocksize in comparison to

existing processing methods for processing stripmap SAR data. The compressed sliding spotlight SAR data in azimuth time slant range domain is provided a reverse bulk RCM correction. Now in the compressed data, each azimuth line corresponds to one distinct beam center Doppler frequency. From the knowledge of this beam center Doppler frequency and slant range, each of the target can be appropriately mapped to ground coordinate system.
PRIOR ART :
Synthetic aperture radar (SAR), conceived by Carl A Willey in 1952, was originally conceptualized in stripmap imaging configuration. The best possible azimuth resolution in this imaging configuration is 1/2 where / is the antenna length along flight direction which is conventionally called as azimuth length. The limitation in the best possible resolution arises out of maximum limit of the length of synthetic antenna and is limited to physical size of the azimuth footprint of the beam. As far as imaging length in flight direction is considered, there is no physical limit imposed by SAR concept.
Almost three decades later another concept of imaging in SAR called spotlight mode emerged. The maximum limit on the length of synthetic aperture is removed by steering the beam towards a particular spot being imaged. Theoretical best possible resolution achievable by spotlight mode is X/4 at infinite observation duration. With practical limits on observation duration, typical resolution obtained by spotlight imaging is

-^lOX. On the flipside, the azimuth length of imaging is restricted within the azimuth footprint of the antenna beam. Primary differences of the above mentioned two imaging modes are summarized in Table-1.
Table-1 Salient Differences in Stripmap and Spotlight Modes

There are certain differences in the SAR processing strategies implemented in these two imaging modes. Two major processing steps range cell migration (RCM) correction and aperture synthesis are generally implemented in different ways. In stripmap case, as the net RCM is small enough for implementation of RCM correction in frequency domain, this correction is usually implemented in Doppler frequency space. Also as the azimuth data length to be processed is much larger than the matched filter length, matched filtering is usually done in frequency domain and image is obtained by inverse transformation in time domain.

In spotlight mode, the RCM encountered is much higher making its correction in frequency domain difficult without phase distortion and the net output image is much smaller than the matched filter length. Consequently RCM correction is employed in time domain, corresponding to center target, usually on receiving. The matched filter is implemented by time domain phase multiplication whereby targets along azimuth direction gets separated in terms of distinct frequency tones, much akin to frequency modulated continuous wave FMCW radar processing. This process is usually termed in SAR literature as deramping. Image is obtained in frequency domain through one forward Fourier transform operation. This peculiar sort of RCM correction leads to best focusing at the center of the image and gradual degradation in focusing as one moves away from image center. They are subsequently sharpened by a secondary range compression operation. Salient differences in the processing strategies of these two modes are presented in Table-2. All the above discussions regarding comparison of these two modes corresponds to one SAR antenna being used for both strip map and higher resolution spot light operation.

During 1994-1996, a new concept of SAR imaging appeared which promised to remove limitations of spotlight mode in achieving larger extent of azimuth stretch of imaging beyond azimuth footprint of the antenna. It is essentially a spotlight SAR with reduced ground speed of the beam trace of the antenna, whereby the aperture time is increased, sufficient enough to achieve improved resolution. This concept is described by D P Belcher and CJ Baker on pages 366-374 of lEE Proc.Radar Sonar Navigation Vol.143, No.6, December 1996.
The processing strategy to be employed for processing of sliding spotlight mode of SAR is described by Riccardo Lanari, Manilo Tesauro, Eugenio Sansoti, Gianfranco Fomaro on pages 1993-2003 of IEEE Trans. On Geosci. and Remote Sensing,Vol. 39, No. 9, Sept. 2001 and on pages 363-372 lEE Proc.-Radar, Sonar Navigation Vol. 148, No. 6, Dec 2001.

The processing of sliding spotlight data has been attempted with following strategy in these disclosures.
1. In order to correct high RCM two step RCM correction is employed.
The first step is bulk RCM correction, identical to that employed in
classical spotlight SAR whereby the RCM of the center target is
compensated in time domain along with a phase compensation operation.
This method converts the sliding spotlight data to stripmap data.
Residual RCM is corrected in the frequency domain during stripmap
processing.
2. Matched filtering is carried out in frequency domain similar to
stripmap processing.
In essence, existing literature has addressed the problem as an extension of classical spotlight processing and the processing exhibits the typical signature of spodight processing with best focusing at the center of the image and loss of focusing as one moves towards image periphery. This loss of focus is also compensated by a secondary compression. This treatment concludes stating that the present analysis is focused on spacebome systems typically characterized by small squint angles during acquisition (often less than 1°) and negligible motion errors. The extension of the proposed method to the airbome case is certainly possible but, in such case, additional and specific issues such as motion compensation and/or high squint angle values in the data acquisition must be carefully considered and may be worth pursing.

As per the above conclusion, from space-bome SARs, operating at typical altitudes of 600 to 800 km it will be difficult to achieve sliding spotlight images larger than the area of the order of 10 kmxlOkm with resolution of the order of 1 m. If one attempts to extend the azimuth extend to 100 km with spacebome sliding spotlight SAR, azimuth squinting in excess of 10° will be required to achieve resolutions of the order of 1 m. As per the above conclusion, present available methods will not be able to process such data.
The difficulties of processing sliding spotlight SAR data for extended azimuth coverage can be best understood from a specific example. Consider a hypothetical spacebome SAR system specified in Table-3 below.

The major handicap in processing this data from this hypothetical spacebome SAR is widely varying Doppler parameters and RCM. From the typical variations in Doppler parameters and RCM, it is evident that histories of individual targets are limited within a Doppler band traced by fore and aft end of the beam. The beam center Doppler will vary between 0 Hz at center of the spot to around ±40000 Hz at azimuth extremities. However, the instantaneous Doppler bandwidth is within 2500 Hz. This fact enables PRF choice determined by the instantaneous Doppler bandwidth. The RCM behaviour shows widely varying linear RCM component to the tune of 4.5 km at the edges of azimuth extent. In real terms, it implies that phase history of one target can get spread over an area extending around 11000 samples along azimuth and 7500 samples along range.
It should be noted that one can process SAR images by time domain correlation by defining matched filter for every individual pixel. But such a brute force processing strategy will be prohibitively expensive. Best computation performance is achieved when block processing is implemented wherein same matched filter is used for compressing all the targets along azimuth within a block. It is needless to mention that matched filter needs to be updated along range as azimuth match filter is a function of range. The block-size is to be chosen such that Doppler parameters are stable within a block. The stability criteria for Doppler rate is --0.003% for achieving optimum focusing and '-0.5% for optimum RCM correction. In all the processor design, block-size is chosen for meeting the Doppler rate criteria optimum for focusing. Such a choice

will yield hardly any block-size as one moves away from spot center. At spot center, this criteria is easily met for considerable block-size. However, as one moves away from spot center, the azimuth size of the block gets smaller.
The upper bound of the azimuth block size is determined by the stability criteria of Doppler parameters governed by RCM correction. Such a criterion will increase the block size in azimuth direction significantly. However, the processing method should be capable of processing the entire azimuth block by a single matched filter fiinction.
OBJECT OF THE INVENTION :
The object of invention is to develop a method for processing an azimuth block of data for sliding spotlight mode of a spacebome SAR which exhibits large drift in Doppler rate parameter within a block with a single matched filter fiinction. This enables achieving maximum block size from the relaxed requirement of stability of Doppler rate parameter for carrying out RCM correction. Consequently the method improves the net computation efficiency of processing.
DESCRIPTION OF INVENTION:
The method for processing spacebome sliding spotiight SAR signal for extended azimuth coverage is carried out in the following steps.

Dividing the data in suitable blocks spread along azimuth and range assuming same Doppler centroid is assumed in the range extent of the block.
Bulk RCM is corrected by shifting the target at range R to R-AR in accordance with the formula
AR= -A/DC?/2- ^/DR_BEAM_CENTRE^^/^ wherein,
/DC andyi)R BEAM CENTRE^-^ ^^^ Doppler centroid and beam center
Doppler rate.
The same bulk RCM correction factor is used for all the range
gates within a block of data.
After bulk RCM correction, a phase compensation is carried by
multiplying the bulk RCM corrected signal with the following
function : exp(-i27i/Dct -infm BEAM CENTRE^^)
The same phase compensation factor is used for all the range
gates within a block of data.
Typically the bulk RCM correction and phase compensation
coefficients will be chosen for the target situated at the center of
the block.
By carrying out steps 2 and 3, the data is converted with a phase
history equivalent to a stripmap SAR with a difference. In the
stripmap SAR all the phase histories of all the targets lying in
the same range gate exhibit same Doppler rate. However, in this
case the phase histories of all the. targets lying in the same range
gate after bulk RCM correction will exhibit different Doppler

rate. Further the target from the target situated at center of azimuth block, the drift in Doppler rate will be more. The net drift in Doppler rate will be least for the block situated at the spot center at which antenna is looking in broadside direction. The drift in Doppler rate will be more if the block is situated away from spot center when antenna is looking at the target with more azimuth squint angle.
A method of block processing of resultant data with linearly drifting Doppler rate along azimuth is carried out in accordance with the method involves resampling of raw data, either in time domain for deramping based processor or in frequency domain for matched filtering in frequency domain. This resampling scheme converts the target histories along azimuth from drifting Doppler rate to constant and identical phase history. This resampling scheme enables the processing of resultant stripmap data with single matched filter frmction.
Resampling scheme in time domain for deramping based processor is carried out as follows.
i. Residual RCM is carried out in frequency domain by taking short burst of data in frequency domain and then bringing back the data to time domain after residual RCM correction.
ii. The raw data is multiplied with matched filter history corresponding to center target of azimuth block.

iii. The resultant phase histories are alligned by convolving with a filter to obtain an image in frequency domain.
iv. Resultant data is resampled to convert phase histories of individual targets from a slow chirp to constant frequency signal.
V. Final image is obtained i frequency domain.
Resampling scheme in frequency domain for frequency domain matched filtering is carried out as follows.
i. The raw data is converted to frequency domain.
ii. Residual RCM is corrected in frequency domain.
iii. The azimuth spectrum is resampled as per resampling
scheme to convert varying phase histories of different targets
to identical phase history for all targets, iv. The time domain history of center target in azimuth block is
converted to frequency domain and identical resampling
procedure is adopted. V. Resampled raw data in frequency domain is multiplied with
complex conjugate of resampled phase history of the center
target, vi. Inversion of the matched filtered signal from frequency to
time domain to obtain image.
5. The compressed image data is subjected to reverse bulk RCM
correction by shifting target at range R to R+AR wherein

AR- -^fY)Qt/2- /iyi)R BEAM CENTRE^^/"^
In the image domain, image points at a particular azimuth time, for all range gates, correspond to one particular beam center Doppler frequency and hence at one azimuth squint angle. From this information, all the targets at particular azimuth time can be repositioned at ground coordinate system.
DESCRIPTION OF FIGURES :
Fig.-l shows a pictorial description of imaging operation of conventional strpmap SAR.
Fig.-2 shows a pictorial description of imaging operation of conventional spotUght SAR.
Fig.-3 shows a pictorial description of imaging operation of sliding spotlight SAR.
Fig.-4 shows a typical behaviour of Doppler graphic representation of the parameters for the hypothetical spacebome sliding spotlight SAR over 100 km azimuth extent of imaging.
Fig.-5 shows a typical behaviour of range cell migration parameters for the hypothetical spacebome sliding spotlight SAR over 100 km azimuth extent of imaging.

Fig.-6 shows a pictorial depiction of translation of Doppler phase history because of bulk RCM correction and phase compensation.
Fig.-7 illustrates range history of targets which appear at the same range gate after bulk RCM correction.
Fig.-8 shows a pictorial depiction of behaviour of phase histories of the targets with drifting Doppler rate during various stages of modified deramping processor scheme incorporating resampling in time domain.
Fig.-9 shows resampling scheme in time domain for deramping base SAR processing.
Fig.-10 shows block schematic of modified reramping based SAR processor incorporating time domain interpolation for processing sliding spot light signal.
Fig.-l 1 shows improvement in impulse response behaviour of the targets spread over 1 second of azimuth flight time at one end of 100 km azimuth extent with conventional and modified deramping processor scheme incorporating time domain resampling.
Fig.-12 shows a pictorial depiction of behaviour of phase histories of the targets with linearly drifting Doppler rate during various stages of

modified frequency domain matched filtering processor scheme incorporating resampling in frequency domain.
Fig.-13 illustrates implementation scheme for implementation in frequency domain matched filter convolution approach to account for linear drift in azimuth chirp rate.
Fig.-14 shows a block diagram of modified matched filter convolution based SAR processor for processing stripmap SAR data with drifting Doppler rate.
Fig.-15 shows an improvement in impulse response behaviour of the targets spread over 1 second of azimuth flight time at one end of 100 km azimuth extent with conventional and modified frequency domain matched filtering processor scheme incorporating frequency domain resampling.
Fig.-16 shows a two dimensional simulation of modified matched filter convolution method with RISAT parameters at 50 km away from spot centre, showing improvement in impulse response (a) with conventional method (b) with modified method.
Fig.-17 shows geometric reference of reverse bulk RCM corrected processed image.

Fig.-18a to 18g shows a two dimensional simulation result for an array of point targets placed at one end of 100 km azimuth extent.
Fig.-19 shows impulse responses of the three targets, shown in Fig. 18(g), after mapping to ground coordinates, the Z-axis is being in dB scale.
Fig.-20 shows relationship of processing filter bandwidth location vis-avis doppler phase history for a set of targets.
(a) Without any error in estimation of azimuth steering rate and doppler centroid of centre target.
(b) With error in estimation of azimuth steering rate and no error in estimation of doppler centroid of centre target.
(c) With a error in estimation of both azimuth steering rate and doppler centroid of centre target.
Fig.-21 shows a demonstration of resilience of the processor developed by the method according to the invention with accurate and erroneous the knowledge of scan rate.
Thus the method according to the invention incorporates processing stripmap SAR data with linearly drifting Doppler rate along azimuth via a resampling process either in time domain (for deramping based processor) or in a frequency domain (for frequency domain matched filtering based processor). The effectiveness of this method is demonstrated by two processing approaches in time as well as frequency domain. The novel resampling scheme ensures that the phase histories of

all the targets along azimuth, lying at a particular range gate after bulk RCM correction, will become identical for all these targets. Thus block processing by a single matched filtering is possible. The increase in block size can be seen in the simulation results presented in Figs. 11 and 15. Possibility of increased block size ensures higher computation efficiency. The azimuth size of the block is govemed by the lower size dictated by the linearity in the drift rate of the Doppler rate or by the net drift in Doppler rate which can be accommodated in residual RCM correction. The range size of the block is govemed by the rate of change of Doppler centroid along range as same Doppler centroid figure is assumed for one block.
Though Bulk RCM correction is a conventional process, in all the existing methods, the bulk RCM correction corresponds to the block center target which has a physical existence. However, in the present method, the bulk RCM correction factor does not correspond to any physical target. It corresponds to trajectory of the beam center line which generates linear Doppler history. This is a novel concept.
Conventional stripmap processing or spotlight processing can be arrived at in the suggested method by very minor modifications in the coding.
Bulk RCM correction and phase compensation scheme used in the method according to the invention is described below.

compensation and bulk RCM correction will drift in proportion to the time difference with respect to the block center target. The drift rate in Doppler rate is increased as one moves away from spot center along azimuth. Subsequent discussion is concentrated on block processing the data where azimuth Doppler rate of the target is drifting linearly. Further, higher order phase component will also change. The block size will get limited by the condition that the drift in Doppler rate is linear and the change in higher order phase term over the block length is not appreciable.
Stripmap processing with drifting Doppler rate scheme used in the method according to the invention is described below:
Block processing of stripmap SAR is possible if Doppler rate is constant or the phase variation due to Doppler rate drift is within tolerable limits of n/4. Beyond this limit the impulse response degrades and this constraint is the only limiting factor in determining the block size. In the following sections an approach is presented for block processing of strip map SAR with drifting Doppler rate. This method will extend the azimuth block size significantly limited by the constraint that linearity of drift rate holds good over the azimuth block extent. Two different implementations are provided. One using deramping method and the other using matched filtering method, are presented for demonstrating the effectiveness of this approach.

^/J^] = ^/J^-l] + A7}4v[^-l])--(25) The resampling is accomplished by equating output and input signals as
Fig." 10 shows the complete processing scheme of sHding spotHght SAR signal using deramping based processor.
Fig." 11 shows effectiveness of this method with one dimensional simulation. For this simulation hypothetical spacebome SAR parameters shown in Table-3 at 50 km away (azimuth) from the spot center and 560 km ground distance from nadir track were assumed. The foj^ MODIFIED =' 715 Hz/sec and /OR RATE ='^'^ Hz/sec were assumed. Conventional
stripmap processing would not have produced sufficient block image there by drastically reducing block processing efficiency. Because of the above mentioned modification, even at the far end, a block data may be processed over 1 second flight duration which is a significant achievement. The duration over which the processor can be applied depends upon/^^ ^^^^ and residual cubic phase term.
In a frequency domain matched filter convolution processing of stripmap SAR data with drifting Doppler rate like the previous example of deramping processing by resampling in time domain, the corresponding processing using frequency domain matched filter convolution is also possible by resampling the signal in frequency domain. The processing methodology is presented in Fig.-12. Let us consider three targets as shown in Fig.-12(a), one at the center of the aperture with beam center at / = 0 and two aft and forward targets with

Fig.-15 shows effectiveness of this method with one dimensional simulation. For this simulation hypothetical spacebome SAR parameters (Table-3) at 50 km away (azimuth) from from spot center and 560 km ground distance from nadir track were assumed. The /^.^ MODIFIED =-715
Hz/sec and /^^^ j^^^ =-6.6 Hz/sec were assumed. Conventional stripmap
processing would not have produced sufficient block image there by drastically reducing block processing efficiency. Because of the above mentioned modification, even at the far end we can process a block data over 1 second flight duration which is a significant achievement. The duration over which the processor can be applied depends upon/^;; ^^^j^
and residual cubic phase term. Fig.-A.16 shows the effectiveness of this method with two dimensional point target simulation.
Geometric distortion correction scheme used in the method according to the invention is described below.
The first step to geometric correction after azimuth compression is reverse bulk RCM correction. After this reverse bulk RCM correction, the image is mapped from Azimuth time - slant range domain to ground coordinate system (x,y) with the origin of reference being kept at nadir location at / = 0 reference. Fig.-17 is drawn with flat earth approximation for ease of illustration of salient points in Geometric correction. The matched filter is generated at t = t^ reference with Doppler Centroid
IDC (^0) • It is to be noted that in the processed image in Azimuth time-

Computer simulation scheme of an array of point targets used in the method according to the invention is given below.
The sliding Spotlight processor has been implemented with matched filter convolution. Raw data has been simulated for the hypothetical SAR configuration shown in Table-3 for a set of point targets with a matrix arrangement of 50mx50m spacing spread over 500mx500m area at a ground range distance of 560 km from nadir track. The targets are placed at 50 km distance, opposite to velocity vector direction, from the spot center of spot size 10km (Range)xlOO km (Azimuth). During raw data simulations, no approximation has been made in computing slant range and hence phase history of the targets as the satellite passes over. From raw data to output ground co-ordinate image stage, all the transformations

in the data are shown in Fig.-18. Impulse responses of the marked targets in Fig.-18(g) is shown in Fig.-19.
One interesting observation needs to be pointed out that the method is resilient to Doppler centroid estimation error and estimation error in pitch rate. In fact the current depiction of simulated data has been processed with Doppler centroid estimation error of 200 Hz and 10% error in the estimation of pitch rate. The reason of resilience of the processor is illustrated in Fig."20. In this figure the signal Doppler History and Processor Bandwidth location is presented for a set of three targets. There is an exaggeration in this figure: in reality, the Doppler histories are more closer and almost overlapped and processor bandwidth is much larger and almost encompassing the signal bandwidth. However, this exaggeration makes the illustration simpler. This explains even in presence of estimation in Doppler centroid and steering error, image will still be processed if the processing filter is located in Doppler phase history. For different errors, the targets will appear at different azimuth time and different slant ranges. However, after geometric correction, the targets will appear at the correct ground coordinates. This effect is illustrated in Fig.-21. This figure shows how the targets are positioned at different azimuth time and slant range after processing and reverse bulk RCM correction for different error in estimation of pitch scanning rates. However, in both cases the targets are correctly located in ground coordinate system after geometric correction.

The method according to the invention, is intended to increase the azimuth extent for sliding spotlight SAR imaging from a space-bome platform. Usually all practical SAR processors are designed with processing over a number of blocks of raw data and consequent seamless stitching of individually processed blocks. As one moves from image center in sliding spotlight imaging, the Doppler parameters and RCM change progressively faster over a block of data. The conventional SAR processing, which requires stability of Doppler parameters over a block, will not lead to appreciable block size as one moves away from image center, leading to inefficient processor architecture. The present method addresses the issue of maximizing the block size even under this condition. The heavy RCM is maximally corrected in this method by bulk RCM correction in time domain. However, this bulk RCM correction does not correspond to a physical target in the image, but to the beam center trace. A corresponding phase compensation converts the data to a stripmap one and the residual RCM is corrected in the stripmap processing. However, the resultant stripmap raw data, so obtained, exhibits the same behaviour of progressively higher drift in Doppler parameters, as one moves away from image center. A modification in matched filtering method is suggested to take care of such drifts within a block as long as the drifts in Doppler rate is linear. The suggested method is one method for interpolation of the resultant stripmap raw data in time domain or frequency domain, depending upon implementation domain of matched filtering. Further, geometric mapping of the final image from azimuth time - slant range reference frame to ground coordinate reference frame has been investigated. The resilience of the method, even

in the presence of Doppler parameter estimation error, to map the image accurately in ground coordinate system has been demonstrated. Apart from analytical development, the method has been demonstrated by processing simulated raw data of a matrix of point targets. For simulation of the raw data, hypothetical SAR parameters (Table-3) were assumed.

WE CLAIM :
A method for processing spacebome sUding spot hght synthetic aperture radar (SAR) signal for extended azimuth coverage comprising the steps of: (i) dividing the data in pre-determined blocks spread along
azimuth and range assuming same Doppler centroid in the
range extent of the block; (ii) converting bulk range cell migration (RCM) in accordance
with the bulk RCM correction scheme described herein
using same bulk RCM correction factor for all the range
gates within a block of data; (iii) compensating the phase by multiplying the bulk RCM
corrected signal in accordance with the phase compensation
scheme described herein using the same phase compensation
factor for all range gates within a bock data; (iv) stripmap processing with drifting Doppler rate in accordance
with the stripmap processing scheme described herein; (v) correcting geometric distortion in accordance with the
geometric distortion correction scheme described herein; and (vi) obtaining an array of point targets using computer simulation
scheme of an array of point targets described herein.

The method as claimed in claim 1, wherein the said stripmap processing step is carried out by deramping based processing of stripmap SAR data with drifting Doppler rate as herein described.
The method as claimed in claim 1, whereien the said stripmap processing step is carried out by frequency domain matched filter convolution processing of stripmap SAR data with drifting Doppler rate as described herein.
A method for processing spacebome sliding spot light synthetic aperture radar (SAR) signal for extended azimuth coverage substantially as hereinabove described and illustrated with reference to the accompanying drawings.

Documents:

076-che-2004-abstract.pdf

076-che-2004-claims filed.pdf

076-che-2004-correspondnece-others.pdf

076-che-2004-correspondnece-po.pdf

076-che-2004-description(complete) filed.pdf

076-che-2004-drawings.pdf

076-che-2004-form 1.pdf

076-che-2004-form 19.pdf

076-che-2004-form 26.pdf

076-che-2004-form 3.pdf

76-che-2004 claims granted.pdf

« Previous Patent

Next Patent »

Patent Number

201926

Indian Patent Application Number

76/CHE/2004

PG Journal Number

05/2007

Publication Date

02-Feb-2007

Grant Date

30-Aug-2006

Date of Filing

30-Jan-2004

Name of Patentee

M/S. INDIAN SPACE RESEARCH ORGANISATION

Applicant Address

ISRO HEADQUARTERS, DEPARTMENT OF SPACE, ANTARIKSH BHAVAN, NEW BEL ROAD, BANGALORE 560 094

Inventors:

#	Inventor's Name	Inventor's Address
1	TRPAN MISRA SCI/ENGR SG	MICROWAVE SENSOR SYSTEM DIVISION, MICROWAVE SENSOR GROUP, SPACE APPLICATIONS CENTRE (ISRO), AHMEDABAD 380 015, GUJARAT

PCT International Classification Number

GO1S13/90

PCT International Application Number

N/A

PCT International Filing date

PCT Conventions:

#	PCT Application Number	Date of Convention	Priority Country
1			NA