Adaptive multi-channel ATI-SAR for moving target detection
MetadataShow full item record
Along-track interferometry (ATI) is a Synthetic Aperture Radar (SAR) technique for moving target detection and estimation of radial velocities. The conventional ATI-SAR involves the acquisition of data from two channels that are separated in the direction of flight path. While the stationary targets share the same signature in the SAR images of these two channels, the moving targets exhibit phase shifts between the two SAR images. Thus, we can identify the moving targets by examining the phase information of the SAR interferogram. For a multi-channel SAR system, SAR interferogram cannot fully exploit the phase information of all channels. In this dissertation, we examine the possibility of generating the moving target indication (MTI) statistic by developing a multi-channel ATI-SAR processing method. The merits of this method are vigorously studied using both simulated data and real data from Multi-Channel Airborne Radar Measurement (MCARM) system. The method we proposed is not feasible without the full calibration of all channels. To deal with this issue, we begin with a simple procedure called global calibration, which is mainly adopted to compensate the physical distance between the two channels in the along-track domain. It is known that the MTI ability of ATI-SAR is substantially affected by the drift angle of the moving platform. As an integral part of this dissertation, we will discuss the effect of drift angle from a STAP point of view. We will also introduce a STAP approach to estimate platform velocity as well as drift angle. The estimation of drift angle does not account for various sources of other errors. To fine calibrate these errors, including which are caused by the drift angle, a two-dimensional (2D) adaptive filtering algorithm called signal subspace processing (SSP) is introduced. The SSP algorithm is applied in the SAR image domain to address the spatially-varying nature of the calibration errors. The SSP algorithm is too time-consuming for a practical radar system. In the final part of this dissertation, we will discuss the parallelization of the proposed multi-channel ATI-SAR processing method and its implementation on high performance computing clusters (HPCC).