Artefact Analysis of Multistatic Three-Dimensional SAR Imaging with Two Linear Trajectories

Daniel Andre, Venkateswaran P. Krishnan, Clifford Nolan · 2025

Read Artefact Analysis of Multistatic Three-Dimensional SAR Imaging with Two Linear Trajectories by Daniel Andre, Venkateswaran P. Krishnan, Clifford Nolan (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-11323.txt.

Introduction

Daniel Andre, Venkateswaran Krishnan, and Clifford Nolan study ghost artefacts in three-dimensional radar imaging. Their paper, from twenty twenty-five, analyzes a multistatic setup. Transmitter and receiver fly separate straight paths over the scene. Such split systems suit satellite swarms and drone teams.

Fine-scale math traces how the flight geometry shapes the image. Some geometries smear false copies of targets into the picture. The authors derive a time-gating rule for the recorded data. Applying the gate keeps the region of interest clean of artefacts. Computer simulations, run under a fixed-time formulation, back the theory. Mission planners can use the rule when choosing flight paths.

Catalog record

CREATORDaniel Andre, Venkateswaran P. Krishnan, Clifford Nolan
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE47 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-11323

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Artefact Analysis of Multistatic Three-Dimensional SAR Imaging with Two Linear Trajectories Daniel Andre, Venkateswaran P. Krishnan, Clifford Nolan Abstract In recent years, radar technology has seen much improvement, making multistatic Synthetic Aperture Radar (SAR) sensing a realistic possibility, for example in satellite constellations or unmanned aircraft systems. With such systems, there then comes the requirement to investigate useful multistatic SAR geometries. We provide a microlocal analysis of a multistatic data acquisition geometry for three-dimensional SAR imaging, where the transmitter and receiver move independently along two straight lines covering a region of interest (ROI). Using microlocal techniques, we show how the data acquisition geometry influences whether artefacts are likely to be present in the resultant image and how they can be avoided. Our main contribution is to provide a time-gating condition on the data which ensures that artefacts are not present in the ROI. As an independent verification, we provide several numerical simulations which follow a time-independent formulation. keywords SAR, Bistatic, Multistatic, Artefact, Singularity, Microlocal Analysis † † equal-contributors: All authors contributed equally to this work 1 Introduction In recent years, much research has been conducted on bistatic radar imaging Cheney and Borden (2009) ; Yarman et al. (2008) ; Ambartsoumian et al. (2018) ; Felea et al. (2016) ; Felea et al. (2022) . The increased interest is partly due to improvements in technology which have made the proposition of multistatic radar sensing systems realisable. Microlocal analysis Grigis and Sjöstrand (1994) ; Hörmander (2003) ; Duistermaat (2011) ; Trèves (1980) has proven to be a powerful tool for understanding how backprojected images of scenes are formed and how image artefacts can appear. Understanding how and where these artefacts may form is vital so that one may have the opportunity to carefully design how the Radar data is collected (the acquisition geometry) in order that such artefacts may be avoided in certain parts of the scene; the so-called regions of interest (ROI). We consider multistatic (multiple bistatic) synthetic aperture radar (SAR) imaging that employs radio wave sources and receivers distributed along a pair of horizontal straight lines ( \gamma_{1},\gamma_{2} ) which are at an angle \beta to each other and these lines are parametrised by s and r respectively. Thus, for N transmitter positions and M receiver positions, we have MN bistatic radar pulse datasets. The SAR geometry is illustrated in Figure 1 . Such a multistatic SAR geometry would provide a virtual two-dimensional SAR aperture, so we expect, in principle, to be able to reconstruct three-dimensional objects, as we have three degrees of freedom ( s,r - the slow time variables and t - the fast time variable). The ability to scan a two-dimensional virtual aperture, whilst constraining the transmitter and receiver (or transceiver) positions to lie only on two single lines could prove beneficial in practical scenarios where the region of space in between the lines may be inaccessible to sensors. It may also be the case that accurately arranging or scanning a transmitter or receiver purely along two straight lines may prove more practical to set-up when compared to accurately scanning a two-dimensional range of transmitter or receiver positions. It is noted that the work here, although directed to sensing with electromagnetic waves, is also relevant to synthetic aperture sensing with other wave modalities including sound waves, which has applications in underwater imaging, medical ultrasound, and seismic waves to image underground structures, used for geological exploration, for example. r s h \gamma_{1}(s) \gamma_{2}(r) (x_{1},x_{2},x_{3}) Scatterer Ground reference plane ( x_{3}=-h ) x_{3}=0 \beta Figure 1: Illustration of the proposed acquisition geometry, showing generic positions of the transmitter \gamma_{1}(s) , receiver \gamma_{2}(r) and a scatterer at x with coordinates (x_{1},x_{2},x_{3}) . There are at least three different ways that the multistatic collection we propose could be obtained in practice. To help in the understanding of the geometry, these are described in the following three cases. Case A: One may have a set of N fixed transmitters distributed over one line, and a set of M fixed receivers distributed over the other. The transmitters would then transmit one pulse each in sequence, with the receivers providing M recordings for each pulse; Case B: As for Case A, but with the N multiple transmitters may be substituted by a single transmitter that flies out along its line, transmitting N pulses sequentially along its aperture, whilst the M receivers record M signals for each of the transmissions; Case C: As for Case B, however the multiple receivers are substituted by a single receiver that flies along its aperture line multiple times. Here, one would arrange the collection so that for each of the N transmitter positions, the transmitter generates a signal M times, whilst the receiver does a full aperture sweep along its line to collecting the M scattered transmitter pulses. For this case, only a single transmitter and a single receiver are required. One of the acquisition geometries considered here was also considered in Felea et al. (2022) , where the potential for extra artefacts to occur in the image was identified as a possibility. These artefacts may arise due to the breakdown of the local condition in the so-called Bölker condition. More specifically, the Bölker condition requires a certain projection map from the wavefront relation of the scattering operator into the cotangent space of the data manifold to be an embedding; please see the discussion at the end of Section 2 for the terminology used here. The latter paper showed that the projection may fail to be an immersion (the local condition). This paper revisits the Bölker condition and shows that by designing the acquisition geometry appropriately (the extent and location of the source and receiver line segments), then it is possible to ensure that the Bölker condition is satisfied so that we can then be confident regarding the absence of the artefacts mentioned in Felea et al. (2022) . In particular, in this paper, we provide a detailed calculation that shows how one may guarantee that the projection map is injective, putting wavefront set elements of the visible scatterers in a one-to-one correspondence with wavefront set elements of the data. We also consider in this paper the situation where the acquistion geometry has the sources and receivers arranged in two non-orthogonal straight line segments, thereby extending previous results, while at the same time showing that the reconstructions are stable with respect to the acquistion geometry. We wish to analyse the image that could be obtained of scatterers, which in practice could include aircraft, buildings, ground terrain, etc., in the ROI (scene). This image is obtained by backprojecting the scattered wave field recorded at the receivers, which we call the data henceforth. We will presently explain what is meant by the process of backprojection but for now, suffice it to say that it is a concept that is omnipresent in almost all imaging methods. For example, in seismic imaging, we see the concept of “migrating” the seismograms to form an image; see Symes (2009) for example. In the radar community, the backprojection image formation is utilized, and has been implemented on the simulated results described in Section 5 Elgy (2019) . We are particularly interested in investigating whether the proposed

Read the complete text