Font Size: a A A

Study On The Improved Methods For The Multichannel SAR Imaging

Posted on:2018-09-03Degree:DoctorType:Dissertation
Country:ChinaCandidate:S S ZuoFull Text:PDF
GTID:1368330542492950Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
Synthetic aperture radar(SAR)is an irreplaceable remote sensor since it works in all-day and all-weather conditions.It is able to obtain the high-resolution and wide-swath observation of the surface form long distance.Because of its powerful function,SAR has attracted more and more attention.SAR is widely used in many areas.In different application,different modes are required,such as spotlight mode SAR,sliding spotlight SAR,stripmap mode SAR and Terrain Observation by Progressive Scans SAR(TOPS SAR).For the single SAR mode,there are many effective imaging methods.However,the united imaging algorithm which is suitable for many SAR modes simultaneously is relatively infrequent in literatures.So it is necessary to develop the united imaging algorithms for the multiple SAR modes and multiple channel SAR.For the conventional single channel SAR,high resolution and wide swath are contradiction to each other.On one hand,the finer the azimuth resolution,the wider the Doppler bandwidth,the greater the pulse repetition frequency(PRF)will be,which is used to avoid the Doppler aliasing,according to the Nyquist criterion.On the other hand,a wider swath requires a lower PRF to avoid the range ambiguity.Therefore,high resolution and wide swath cannot be obtained in a single channel SAR(SCSAR)system at the same time.To deal with this problem,azimuth multichannel SAR(MCSAR)has been proposed in the literature.MCHSAR has more degrees of freedom(DOF)than SCSAR.In an MCHSAR system,the time sampling is compensated by the space sampling,since there are many channels in azimuth direction and the minimum antenna area constraint is relaxed.As a result,MCSAR may provide HRWS imaging.The dissertation focuses on the MCSAR and multiple SAR modes imaging.The main content is as follows.1.For high-resolution wide-swath(HRWS)synthetic aperture radar(SAR)imaging algorithms,signal reconstruction is a key step.The steering vector plays an important role in signal reconstruction,which can be constructed by the ambiguity components.The information of the ambiguity components,e.g.,number and index,is usually regarded to be constant and known.However,we find that the information of ambiguity components is always a piecewise function of the baseband frequency.This means that the steering vector cannot be pre-constructed accurately and it will negatively affect the signal reconstruction.This paper presents an improved signal reconstruction method based on Doppler spectrum estimation.The proposed method can estimate the variant ambiguity components to form the steering vector exactly by Capon estimation.As a result,the method is able to restore the Doppler spectrum entirely and performs well on noise reduction.Moreover the baseband Doppler centroid and antenna pattern can be obtained in the proposed method.Simulated data and airborne raw data are processed to validate the algorithm.2.In the high-resolution and wide-swath synthetic aperture radar imaging,channel error calibration is an important step.The key to the channel error calibration methods based on the subspace is accurate division of the subspace.This paper analyses the variant of the ambiguity components with the baseband frequency and proposes an improved channel error calibration method for the azimuth multichannel SAR.The proposed method can get the accurate information of the ambiguity components by the spatial spectrum estimation and calibrate the channel error precisely.The real data results validate the method.3.For the clutter suppression in the MCSAR system,the virtual multiple channel clutter suppression is proposed.It is a key step for the multiple channels SAR-GMTI system.However because of the Doppler spectrum aliasing,conventional DPCA methods for clutter suppression are ineffective.For the clutter suppression methods based on the STAP,their performance is limited by the channel number of the SAR system.To solve the problems mentioned above,this paper proposed a clutter suppression method based on the virtual multiple channels.In the method,moving target Doppler spectrum compression technique is applied to the multiple channel SAR echoes.With the help of the moving target Doppler spectrum compression technique,the total number of the ambiguity components of MCSAR system is decreased.Then the proposed Doppler spectrum diagram deramp method is used to flatten the slant Doppler spectrum diagram of the MCSAR signal.Next the channel number is increased by the virtual multiple channel processing.At last the optimum weight vector is constructed to suppress the clutter.In the proposed method the number of the Doppler ambiguity components keeps constant but the number of channels is rising.Due to the increase of the channel number,there is more freedom degree to constraint the side lobes of the antenna pattern.The proposed method has a good performance in clutter suppression meanwhile it is able to suppress the false moving target.The real data results validate the method.4.For the single channel and multiple channel SAR imaging,a new sub-aperture rectangular-coordinate-system(RCS)fast back projection algorithm based on image spectrum compression is proposed.The fast back projection algorithm(FBP)and fast factorized back projection algorithm(FFBP)are both based on the local polar coordinate system(LPCS).In these algorithms,sub-aperture data are firstly processed to form a sub-image in LPCS.Then,sub-images in different LPCS are fused to an image in a new LPCS.Since the azimuth pixel intervals of sub-images are coarse,the computational efficiency to obtain a sub-image is high.Therefore,it is efficient to process the raw data as faster as using the traditional algorithm in the range-Doppler domain.However,at the stage of sub-images fusion,sub-images in several different LPCS are needed to project to a new LPCS,the relationship between these LPCS should be utilized to process the sub-images fusion pixel by pixel.All of these lower the computational efficiency or the processing precision by adopting some approximations.In this paper,a new sub-aperture rectangular-coordinate-system(RCS)fast back projection algorithm based on image spectrum compression is proposed.In this method,the sub-images are formed in local RCS and sub-images fusion is simple by using coordinate translation.The method is also extended to raw data of several SAR modes and muiltiple channel SAR.The simulation data is processed to validate the proposed method.
Keywords/Search Tags:Multichannel synthetic aperture radar (MCSAR), high-resolution and wide-swath (HRWS), Multichannel signal reconstruction, channel error calibration, clutter suppression, multiple SAR modes imaging, back projection algorithm
PDF Full Text Request
Related items