Moving Target Detection and Trajectory Estimation Using SAR Data

Conventional Synthetic Aperture Radar processing leads to moving targets being imaged displaced from their real positions and blurred. This is mainly due to the used reference phase no longer being valid, including an additional Doppler-shift and a Doppler-spread. Since these effects are not considered in the focusing algorithm, the resulting SAR image shows the moving targets defocused and/or at wrong positions, depending on the motion direction. Basically, the Doppler-shift causes a misplacement of the moving targets in the azimuth direction, and the Doppler-spread causes their defocusing in the resulting image. This paper presents some state of the art algorithms that enable the focusing, repositioning, and trajectory estimation of the moving targets. 1.0 INTRODUCTION Conventional Synthetic Aperture Radar (SAR) processing leads to moving targets being imaged displaced from their real positions and blurred [1]. This is mainly due to the used reference phase no longer being valid, including an additional Doppler-shift and a Doppler-spread. Since these effects are not considered in the focusing algorithm, the resulting SAR image shows the moving targets defocused and/or at wrong positions, depending on the motion direction. Basically, the Doppler-shift causes a misplacement of the moving targets in the azimuth direction, and the Doppler-spread causes their defocusing in the resulting image. The development of multi-channel SAR systems facilitates implementation of sophisticated techniques for the detection and further processing of moving objects [2], [3]. It is well known that moving objects induce a variation on the Doppler history, that can be approximated as a quadratic function for low squint angles [4]; the variation of quadratic and linear terms causes, respectively, an azimuth smearing of the energy and an azimuth displacement of the peak, due to the mismatching of the reference function and the wrong range migration compensation. The main difficulty is that the two velocity components are correlated, and the assumption of their independence is not correct, as will be shown in the next section. The processing techniques for moving target processing can be divided into multi-channel and singlechannel. For the azimuth velocity estimation in multi-channel SAR systems, typical approaches consist in using Doppler rate filters to select the velocity that better focuses the target; for the range velocity estimation the techniques are based on the Doppler shift induced by the moving targets, as the Along Track Interferometry (ATI) [5]. Another popular type of estimation algorithm is called Space Time Adaptive Processing algorithms (STAP) [6] that requires an array of antennas. Moving Target Detection and Trajectory Estimation Using SAR Data 7 2 STO-EN-SET-235 Although multi-channel SAR systems are currently available, both airborne and space-borne, the raw data acquired by them is usually of restricted access. For the open scientific community only single channel data is typically made available, a context where moving target processing is more difficult. This paper summarizes some state of the art algorithms published in recent literature that enable the focusing, repositioning, and trajectory estimation of moving targets using single channel SAR data. It starts by deducing the effects induced by a moving target on the SAR image algorithms. Then, it considers the modifications necessary to the wavefront reconstruction algorithm in order to correctly image the moving targets. A well-known limitation, the blind angle ambiguity, is addressed followed by the possibility of using the amplitude information, contained on the received echoes, to solve it. Section 3 presents a complete framework and summarizes several techniques recently published for moving targets processing using single channel SAR, such as digitally spotlighting the signatures, developing an ambiguity function for ground moving target indication (GMTI) and algorithms for their trajectory estimation. Section 4 draws the conclusions. 2.0 MOVING TARGETS PROCESSING This section deduces the effects induced by moving objects on the received SAR signal, and present the modifications necessary to the wavefront reconstruction algorithm [7] in order to obtain focused. The blind angle ambiguity limitation is also addressed. 2.1 Focusing Moving Targets Let us consider the scenario presented in Figure 1, where, for simplicity, a single moving target is present. The radar platform travels at speed vvRR illuminating a moving point-like target with velocity �vvxx ,vvyy� = (μμvvRR , bbvvRR) defined in the slant-plane (xx,yy). Pair (μμ, bb) = ( vvxx vvRR , vvyy vvRR ) is the target relative velocity with respect to the radar.