Synthetic aperture radar imaging exploiting multiple scattering

Abstract

In this paper, we consider an imaging scenario, where a bi-static synthetic aperture radar (SAR) system is used in a multiple scattering environment. We consider a ray-theoretic approximation to the Green's function to model a multiple scattering environment. This allows us to incorporate the multiple paths followed by the transmitted signal, thereby providing different views of the object to be imaged. However, the received signal from the multiple paths and additive thermal noise may interfere and produce artifacts when standard backprojection-based reconstruction algorithms are used. We use microlocal analysis in a statistical setting to develop a novel filtered-backprojection type image reconstruction method that not only exploits the multi-paths leading to enhancement of the reconstructed image but also suppresses the artifacts due to interference. We assume a priori knowledge of the secondorder statistics of the target and noise to suppress the artifacts due to interference in a mean-square error sense. We present numerical simulations to demonstrate the performance of our image reconstruction method. While the focus of this paper is on radar applications, our image formation method is also applicable to other problems arising in fields such as acoustic, geophysical and medical imaging.

Moving Target Artifacts in Bistatic Synthetic Aperture Radar Images

“…(iv) The analysis is based on microlocal analysis [16]. Therefore, the underlying methodology is general and can be used to analyze moving target positioning errors in other SAR modalities [1]- [4], [8]- [11], [40].…”

“…Therefore, assuming a constant velocity v h for the entire scene, we form a reflectivity image of the scene with the following backprojection operator: (8) whereρ v h (z) is the reconstructed image corresponding to the hypothetical velocity v h and Q v h is a filter that can be determined with respect to a variety of criterion [13], [46], [47]. We assume that Q v h satisfies an assumption similar to the one in (7).…”

“…The received signal is generated using (4). The images are reconstructed using the filtered-backprojection inversion given by (8) under the assumption that the scene is stationary. All of the reconstructed images are shown in dB scale.…”

Improved detection probability of a radar target in the presence of multipath with prior knowledge of the environment

“…Further, with the development of advanced and computationally efficient EM tools, EM propagation models and simulations are being incorporated into radar and sensing problems as well, for example, in [10,11]. Recently, multipath exploitation has been receiving interest in radar problems in order to achieve more robust performance and degrees of freedom in detection and estimation problems [25][26][27][28][29][30]. Along these lines, this work investigates the problem of target detection in a multipath environment.…”

Improved radar target time-delay estimation with multipath exploitation

“…Recently however, our view on multipath is changing: with the proper knowledge of the environment, multipath may be considered to be an opportunity to obtain additional degrees of freedom by providing virtual radar sensors that observe the target from different aspects, [5][6][7][8][9][10]. Although there is strong evidence that one can improve detection and estimation performance in radar through multipath exploitation, the amount of improvement still depends on the radar-target geometry, the corresponding multipath structure, signal bandwidth, and the available signal noise ratio (SNR) levels [11][12][13].…”