Research On Sparse Processing Method For Airborne/spaceborne GMTI Radar System

Airborne/spaceborne ground moving target indication (GMTI) radar system playsan important role in military applications. Since airborne/spaceborne radar usuallyworks in the downward-looking status thus causing a spreading clutter Dopplerbandwidth because of the high speed of the platform, the moving targets may besubmerged in clutter. At present, space time adapting processing (STAP) andmulti-channel SAR/GMTI techniques are the two fundamental clutter suppressionmethods for high-speed platform radar. In practical applications, it is hard to acquirethe sufficient Independent and Identically Distributed (IID) sample supports toestimate the covariance matrix due to the heterogeneity of the clutter, in the endcausing a dramatic degreadation of the STAP processor performance. For themulti-channel SAR/GMTI system, the increasing demands for high resolution of SARimages and wide swath coverage have resulted in a huge SAR raw data, which posesmajor constraints in the operation of SAR to transmit the data to ground, or to storethem onboard. In this thesis, novel GMTI methods with limited samples are studied.The main contributions of this thesis are as follows:1、For conventional Space-Time Adaptive Processing (STAP), sufficient IIDsample supports are hard to obtain. To mitigate this problem, a Compressive Sensing(CS)-based Ground Moving Target Indication (GMTI) method is proposed. In theproposed method, firstly the space-time data of the interested range bin is transformedto spatial-temporal frequency fields to accumulate the signal energy and thecorresponding redundant dictionary is constructed; Then several primary clutterspectrum peaks are extracted by Bayesian compressive sensing technique to estimatethe clutter ridge; Finally, the weighted l1minimization optimization model is used torealize the ground moving target indication(GMTI) without estimating the covariancematrix. Simulation results demonstrate the excellent GMTI performance of theproposed method.2、In the conventional SAR ground moving targets indication (GMTI) method, thesample number is heavily large, which greatly increase the data transmission andstorage load. To mitigate this problem, a SAR/GMTI method based on compressivesensing is proposed in this paper. In the proposed method, compressive sensing theoryis used to perform the azimuth direction focus with sparse sampled raw data. Theconventional displaced phase center antenna (DPCA) technique is adopted to suppress clutter. Theoretical analysis shows that the proposed method can be applied to cluttersuppression of the sparse sampling data of dual channels and the influences of motionparameters (range velocity/acceleration, azimuth velocity) on target imaging areanalysed in detail. The results of simulated and real data processing verify that theproposed method has excellent clutter suppression performance in the case of highsignal to clutter ratios (SCNR).3、For the SAR-based GMTI system, the clutter scattering centers of surveillancearea are usually non-sparse. In this case, the sampled raw data of single channel cannot be significantly reduced. Considering the fact that the correlations among themulti-channel SAR images are high, a SAR/GMTI method based on jointly sparse andweighted l1optimization model is proposed. In the proposed method, dual channelSAR raw data are jointly processed. Firstly, a transform matrix is constructed toseparate the energy support areas of moving targets from that of all scattering centers,and then we convert dual-channel SAR imaging to single-channel imaging and movingtargets reconstruction. Then, we can roughly obtain the energy support areas of allscattering centers via CS. Finally, based on the acquired energy support areas above,clutter suppression and GMTI is achieved by solving a weighted l1optimization model.The proposed method is applied to the SAR raw data sparsely sampled in1-dimensiondomain (azimuth) and in2-dimension domains (range and azimuth), respectively.Simulated and real data experiments demonstrate that the proposed method performswell with sparse sampled raw data, even if clutter scattering centers have a low sparselevel.4、For the cases of clutter scattering centers are non-sparse,we propose a cluttersuppression and GMTI method with sparse sampled data for dual-channel SAR. In theproposed method, one channel periodically transmits and receives pulses at Nyquistsampling rate, and other channels sparely and randomly receive pulses. Firstly, rawdata of the channel with full sampling are used to perform SAR imaging. Thenutilizing the acquired SAR image above to be prior-knowledge, the clutter included inother channels data is suppressed. Since the moving targets are sparse in space afterclutter suppression, the moving targets image can be accurately recovered bycompressive sensing. The proposed method is robust to amplitude and phase errorsbetween the channels. The experiment results demonstrate that the static and dynamicinformation loss of the interest area is slight, even if the sampled data are significantlyreduced. 5、The conventional ScanSAR imaging method acquires wide swath coverage at acost of severe azimuth resolution loss. Moreover, scalloping is also a major ScanSARdrawback. To overcome the aforementioned problem, a compressing sensing-basedimaging and GMTI method is proposed for ScanSAR mode. In the proposed method,CS is utilized to perform azimuth imaging by processing jointly sparse aperture data ofeach subswath after range compression and range cell migration correction (RCMC).The proposed method has high azimuth resolution. Moreover, it can overcomescalloping problem without resolution loss. An excellent clutter suppressionperformance using the proposed method can be achieved in the application ofmulti-channel SAR raw data processing.