Synthetic Aperture Radar (SAR) images of moving targets are often displaced and defocused, making reliable localisation and recognition difficult. The problem becomes more pronounced when target behaviour departs from simple motion assumptions and is better described by more complex kinematics, such as six-degrees-of-freedom (6-DoF) motion. Although many autofocus methods rely on subaperture or full-aperture formulations, such approaches can become restrictive when motion varies nonlinearly over the aperture and finer compensation is required. Multistatic SAR provides additional receiver diversity and, when combined with frequency and polarimetric diversity, it helps constrain this problem; however, such information is not usually exploited explicitly in pulse-by-pulse autofocus. This paper introduces a multistatic, frequency- and polarimetry-adaptive autofocus algorithm (F-Pol) that extends Localised Threshold Sharpness (LTS) into the joint frequency-polarimetric domain by encouraging agreement across frequency sub-bands and polarisation channels. Experimental validation using the Cranfield University Ground-Based SAR system with emulated target motion shows that, across the tested signal-to-noise ratio (SNR) range of -5 to 20 dB, F-Pol gives improved image quality and target motion estimation relative to Intensity-Squared (ISQ) and LTS. Laboratory and supplementary simulation results indicate that joint frequency and polarimetric diversity enhances multistatic SAR autofocus performance.

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 transmitters and receivers effectively constitute two linear arrays, which survey 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.

The ability to control sidelobes in a SAR image is critical to forming images that are useful for interpretation and exploitation. QinetiQ has developed the RIBI sensing system, which utilises a distributed coherent array of sensors to produce multistatic images. These systems require techniques from outside the traditional radar domain to utilise the theoretical resolution possible in synthesising a coherent aperture from multiple disparate collections. This paper develops previously published work on using compressive sensing techniques to suppress sidelobes in SAR images to develop a higher-fidelity measurement model. Using Cranfield University's GBSAR System a series, experimental measurements are conducted, and image estimation techniques are applied to this real data. It demonstrates an improvement in recovery performance over an isotropic measurement matrix, and discusses areas which require further development.

Synthetic Aperture Radar (SAR) plays a vital role in the surveillance of terrestrial and maritime targets, which are commonly in motion. As such, the ability to perform accurate real-time focusing and localisation on moving targets, particularly those moving with complex motion, is desired. Many existing autofocus algorithms struggle to achieve this and rely on sub-aperture processing of SAR data to estimate and compensate for phase errors attributed to unknown target motion. This paper presents a new metric-based autofocus approach, called Localised Threshold Sharpness (LTS), which employs multistatic SAR data to localise and focus a target moving with up to six degrees of freedom motion on a real-time, pulse-by-pulse basis. The algorithm is verified with experimental data, and its performance is compared against the performance of an existing measure of image sharpness suitable for pulse-by-pulse autofocusing, namely the intensity-squared metric, with varying levels of added noise. Normalised cross-correlation results demonstrate a resemblance of at least 80% between Multistatic SAR images focused via LTS autofocus and Multistatic SAR images ideally focused using target motion knowledge for signal-to-noise ratios above 3 dB.

There is significant interest in multistatic SAR image formation, due to the increased development of satellite constellations and UAV swarms for remote sensing applications. The exploitation of the finer resolution and wider coverage of these geometries has been shown to reduce the often-impractical data collection requirements of 3D SAR imagery; this offers advantages such as improved target identification and the removal of layover artefacts. This paper presents a novel polarimetric generalisation of the SSARVI algorithm, which was previously developed to exploit sparse aperture multistatic collections for 3D SAR image formation. The new algorithm presented here, named the PolSSARVI algorithm, combines polarimetrically weighted interferograms for determining the 3D scatterer locations from sparse aperture polarimetric collections. The bistatic generalised Huynen fork polarimetric parameters are then determined for the multistatic PolSSARVI 3D SAR renderings. This new approach was tested on both simulated and experimental data. Experimental imagery was formed using measurements from the Cranfield GBSAR laboratory.

Synthetic Aperture Radar (SAR) is a side looking remote sensing radar imaging mode, and due to the SAR image formation process being based on the assumption of a stationary scene, target motion causes signature defocusing and displacements. In particular, targets with motion towards the radar, i.e. moving at constant speed in a direction towards and perpendicular to a monostatic radar trajectory, appear both focused and displaced in azimuth in corresponding SAR images. We propose the use of a reverse-path multistatic SAR configuration, consisting of one co-located transmitter/receiver travelling with azimuth angular velocity omega, and a second receiver travelling in the opposite direction with angular velocity -3omega, at a similar grazing angle and with the centre of each SAR aperture approximately co-located. The monostatic and bistatic apertures then cover close to the same extent in K-space, but have filled it in a different azimuthal direction from one-another during the simultaneous collections. This results in a pair of images which have moving target displacements in opposite directions, and which should also be suited to coherent subtraction of ground clutter, allowing detection together with SAR resolution imaging of range-moving targets against strong background clutter. We present results, both simulated and measured in the Cranfield University ground-based SAR laboratory, showing the effectiveness of the method in detecting and imaging of moving targets against strong clutter.

Through-wall synthetic aperture radar (SAR) imaging is of significant interest for security purposes, in particular when using multi-static SAR systems consisting of multiple distributed radar transmitters and receivers to improve resolution and the ability to recognise objects. Yet there is a significant challenge in forming focused, useful images due to multiple scattering effects through walls, whereas standard SAR imaging has an inherent single scattering assumption. This may be exacerbated with multi-static collections, since different scattering events will be observed from each angle and the data may not coherently combine well in a naive manner. To overcome this, we propose an image formation method which resolves full-wave effects through an approximately known wall or other arbitrary obstacle, which itself has some unknown "nuisance" parameters that are determined as part of the reconstruction to provide well focused images. The method is more flexible and realistic than existing methods which treat a single wall as a flat layered medium, whilst being significantly computationally cheaper than full-wave methods, strongly motivated by practical considerations for through-wall SAR.

AbstractSatellites and drone swarms can be used to collect multistatic Synthetic Aperture Radar (SAR) images. Synthetic Aperture Radar images can be used for Intelligence Surveillance and Reconnaissance. One method is to use Coherent Change Detection (CCD) to identify changes such as objects or tracks in the scene. This paper investigates a two‐stage change detector, formed using intensity change and CCD images, extended to laboratory measured multistatic SAR data. A variety of performance metrics are used to quantitatively assess the results. Bistatic results are compared to a variety of multistatic and fully polarimetric results. The improvement in performance of multistatic and fully polarimetric images over bistatic images is shown. Additionally challenges and limitations of using multistatic datasets are highlighted.

Synthetic Aperture Radar (SAR) Coherent Change Detection allows for the detection of very small scene changes. This is particularly useful for Intelligence, Surveillance and Reconnaissance as small changes such as vehicle tracks can be identified. Rapidly collecting repeat pass SAR imagery is important in these applications. For space-borne platforms, such repeat passes may however have significant differences, or baselines. Coherent Change Detection products are reliant on high coherence for good interpretability. This work investigates the sources and levels of incoherence associated with bistatic SAR imagery for a variety of baselines using simulations and measured laboratory data for two ground types. Additionally, spatially variant incoherence trimming is implemented. The paper shows the importance of angle-dependant backscatter on the coherence of sub-resolution cell scatterers.

AbstractWith the advent of constellations of SAR satellites, and the possibility of swarms of SAR UAV's, there is increased interest in multistatic SAR image formation. This may provide advantages including allowing three‐dimensional image formation free of clutter overlay; the coherent combination of bistatic SAR geometries for improved image resolution; and the collection of additional scattering information, including polarimetric. The polarimetric collection may provide useful target information, such as its orientation, polarisability, or number of interactions with the radar signal; distributed receivers would be more likely to capture any bright specular responses from targets in the scene, making target outlines distinct. Highlight results from multistatic polarimetric SAR experiments at the Cranfield University GBSAR laboratory are presented, illustrating the utility of the approach for fully sampled 3D SAR image formation, and for sparse aperture SAR 3D point‐cloud generation with a newly developed volumetric multistatic interferometry algorithm.

Synthetic Aperture Radar (SAR) Coherent Change Detection (CCD) allows for the detection of very small scene changes. This is particularly useful for reconnaissance and surveillance as small changes such as vehicle tracks can be identified. In some situations, it is desirable to rapidly collect repeat pass SAR images for use in change detection, and multistatic geometries may facilitate this. Such repeat passes may however have significant baselines, particularly for satellite-based platforms, though CCD products are reliant on high coherence for good interpretability. This work investigates the sources and levels of incoherence associated with bistatic SAR imagery with increasing baselines using simulations and measured laboratory data.

Synthetic Aperture Radar (SAR) renderings in 3D provide additional target information when compared to 2D by separating out features overlaid in height. However, the required 2D SAR aperture, when Nyquist sampled, necessitates large scanning times that would be impractical for most realistic collections. This research has developed a novel volumetric approach to sparse aperture 3D SAR imaging, which is applicable to bistatic SAR near-field geometries, a generalization of far-field cases. This approach is first demonstrated in simulation and then applied to a measured scene containing a model vehicle target, producing sub-Nyquist sampled 3D SAR renderings.

Multirotor Unmanned Air Systems (UAS) represent a significant improvement in capability for Synthetic Aperture Radar (SAR) imaging when compared to traditional, fixed-wing, platforms. In particular, a swarm of UAS can generate significant measurement diversity through variation of spatial and frequency collections across an array of sensors. In such imaging schemes, the image formation step is challenging due to strong extended sidelobe; however, were this to be effectively managed, a dramatic increase in image quality is theoretically possible. Since 2015, QinetiQ have developed the RIBI system, which uses multiple UAS to perform short-range multistatic collections, and this requires novel near-field processing to mitigate the high sidelobes observed and form actionable imagery. This paper applies a number of algorithms to assess image reconstruction of simulated near-field multistatic SAR with an aim to suppress sidelobes observed in the RIBI system, investigating techniques including traditional SAR processing, regularised linear regression, compressive sensing. In these simulations presented, Elastic net, Orthogonal Matched Pursuit, and Iterative Hard Thresholding all show the ability to suppress sidelobes while preserving accuracy of scatterer RCS. This has also lead to a novel processing approach for reconstructing SAR images based on the observed Elastic net and Iterative Hard Thresholding performance, mitigating weaknesses to generate an improved combined approach. The relative strengths and weaknesses of the algorithms are discussed, as well as their application to more complex real-world imagery.

There is great interest in multistatic synthetic aperture radar (SAR) systems as they are capable of providing high resolution images. These systems could prove promising candidates for provision of surveillance for both military and civilian interest. Both multistatic SAR and its counterpart, multistatic inverse synthetic aperture radar (ISAR), are limited by their assumptions of observing a stationary target from a moving platform and vice-versa. Hence, without adequate target motion compensation, their resultant radar images appear defocused. Arranging experiments capable of providing repeatable multistatic hybrid SAR/ISAR data of real moving targets can be difficult and costly. One viable approach is the novel method presented in this study, whereby multistatic hybrid SAR/ISAR data can be collected of a target moving with a theoretical motion, without the requirement of an actual moving target – the theoretical motion is brought about through the appropriate motion of antennas. The study demonstrates, both through simulation and experimentation, how radar trajectories of a given SAR system can be altered to arrive at the equivalent setup of observing a moving target. Results from simulation and from an experiment conducted at the Cranfield University Ground-Based SAR (GBSAR) laboratory are presented, showing the utility of this approach.

Synthetic Aperture Radar (SAR) Coherent Change Detection (CCD) allows the detection of very small scene changes but is typically reliant on a high degree of similarity in the radar trajectories, with a small baseline. In the case of multistatic SAR imagery, such as those formed by a constellation of SAR satellites, the radar trajectories may have a greater baseline than those collected by a monostatic system such as an aircraft. This paper investigates the effects of multistatic trajectories on the measured coherence between imagery, and how this relates to the spatial frequency (K-space). In particular, the case where radar platform trajectories are greatly dissimilar, but where the K-space image supports still contains a high degree of overlap, is investigated. This paper uses multistatic SAR collections measured at the Ground Based SAR Laboratory at Cranfield University.

X-ray backscatter imaging is a powerful technique for medical, aerospace, and security applications. Conventionally, a pinhole is commonly used for focusing X-ray, but there is always a desire to enhance the signal-to-noise-ratio (SNR) and optical throughput compared to a single pinhole. The main aim of this paper is to present a new X-ray backscatter imaging system which was inspired by a Twisted Slit collimator system called the Vortex Collimator and compare the optical throughput and the imaging performance with that of the Twisted Slit' collimator [G. Jaenisch et al., "Scatter imaging - simulation of aperture focusing by deconvolution," (IEEE, Piscataway, NJ, 2017), p.301-306; G. Jaenisch, S. Kolkoori & C. Bellon, "Quantitative simulation of back scatter X-ray imaging and comparison to experiments," 1-11 (2016)] and the Pinhole imaging systems for axial point sources, where the pinhole system was used purely for comparison purposes. All the comparisons were performed through Ray tracing (TracePro) simulation software. This work shows that the Vortex design yields similar to 4% higher SNR/optical throughput than that of the Twisted Slit collimator, and similar to 42.5% higher transmittance. Furthermore, the opening of the Vortex Collimator was increased and reduced to observe the performance, resulting in about similar to 1% transmittance increment when the opening was increased. Also, thicknesses of the Vortex Collimator and Twisted Slit collimator were increased and reduced and found that reducing the thickness appears to increase the system's throughput marginally. (C) 2022 Society for Imaging Science and Technology.

With the advent of constellations of SAR satellites, and the possibility of swarms of SAR UAV's, there is increased interest in multistatic SAR image formation. This may provide advantages including allowing three-dimensional image formation free of clutter overlay; the coherent combination of bistatic SAR geometries for improved image resolution; the collection of additional scattering information, including polarimetric. The polarimetric collection may provide useful target information, such as its orientation, polarizability or number of interactions with the radar signal; distributed receivers would be more likely to capture any bright specular responses from targets in the scene, making target outlines distinct. Highlight results from multistatic polarimetric SAR experiments at the Cranfield University GBSAR laboratory are presented, illustrating the utility of the approach.

It is advantageous to produce Synthetic Aperture Radar (SAR) renderings in three dimensions as these allow the separation of features in height, providing additional target information. A Nyquist sampled 2D SAR aperture produces high quality 3D imagery, however the large scanning time and data storage requirements make this method impractical for use in many scenarios. This paper investigates the formation of 3D target renderings from sparsely sampled 2D aperture Multistatic SAR geometries. The investigation employed both simulations and measured multistatic data collected at the Ground Based SAR (GBSAR) Laboratory at Cranfield University.

Synthetic Aperture Radar (SAR) Coherent Change Detection (CCD) allows the detection of very small scene changes, such as from ground subsidence or vehicle tracks, with applications both civilian and military. This high sensitivity to small changes can mean that differences in collection geometry or in polarisation can lead to significant changes in image coherence. This paper investigates the coherence between different bistatic SAR image geometries and their corresponding spatial frequency supports. It also investigates methods used to model coherence. The investigation employed both simulations and measured multistatic data collected at the Ground Based SAR Laboratory at Cranfield University.