Networked sensing refers to the capability of multiple wireless terminals to cooperate with the aim of enhancing specific figures of merit, e.g., positioning accuracy or imaging resolution. Regarding radio-based sensing, it is essential to understand when and how sensing terminals should cooperate, namely the best strategy that trades between performance and cost (e.g., energy consumption, communication overhead, and complexity). This tutorial paper revises networked sensing from a wavefield interaction perspective, aiming to provide a general theoretical benchmark to evaluate its imaging performance bounds and to guide the sensing cooperation accordingly. Diffraction tomography theory (DTT) is the method to quantify the imaging resolution of any radio sensing experiment from inspection of its spectral (or wavenumber) content. In networked sensing, the image formation is based on the back-projection integral, valid for any network topology and physical configuration of the terminals. The wavefield networked sensing is a framework in which multiple sensing terminals cooperate during the acquisition process to maximize the imaging quality (resolution and sidelobes suppression) by pursuing the wavenumber tessellation principle. We discuss all the coherent data fusion possibilities between sensing terminals and possible killer applications. Remarkably, we show the possibility that the proposed method allows obtaining high-quality images of the environment in limited bandwidth conditions, leveraging the coherent combination of multiple multi-static low-resolution images.

The main interest in using synthetic aperture radar (SAR) technology in automotive scenarios is that arbitrarily long arrays can be synthesized by exploiting the natural motion of the ego vehicle, enabling finer azimuth resolution and improved detection. All of this is achieved without increasing the hardware complexity in terms of the number of physical antennas. In this paper, we start by discussing the application of SAR imaging in the automotive environment from both theoretical and experimental perspectives. We proceed by describing an efficient processing workflow and we derive the rough number of operations required to focus an image proving the real-time imaging capability of the system. The experimental results are based on open road data acquired using an eight-channel radar at 77 GHz, considering side-looking SAR and forward SAR. The results confirm the idea that SAR imaging can be successfully and routinely used for high-resolution mapping of urban environments in the near future.

Deep learning solutions have recently demonstrated remarkable performance in phase unwrapping by approaching the problem as a semantic segmentation task. However, these solutions lack explainability and robustness to unseen conditions, and they often need a large amount of data for training. By contrast, traditional phase unwrapping algorithms, such as PUMA, rely on principled pipelines that estimate the phase through optimization solvers, despite often failing under severe noise conditions. In this work, we show how to exploit the benefits of both approaches by proposing a way to combine deep neural networks with iterative energy minimization algorithms based on graph cuts. We implement a differentiable version of the PUMA algorithm, NeuralPUMA, which we integrate into a traditional deep learning pipeline to implicitly learn to preprocess the wrapped phase into an intermediate representation that improves the algorithm solution. Through extensive experiments, we show that our approach effectively improves the performance of PUMA in noisy conditions and outperforms recent deep learning methods, while also requiring less training data and simpler neural architectures.

Coherent multistatic radio imaging represents a pivotal opportunity for forthcoming wireless networks, which involves distributed nodes cooperating to achieve accurate sensing resolution and robustness. This paper delves into cooperative coherent imaging for vehicular radar networks. Herein, multiple radar-equipped vehicles cooperate to improve collective sensing capabilities and address the fundamental issue of distinguishing weak targets in close proximity to strong ones, a critical challenge for vulnerable road users protection. We prove the significant benefits of cooperative coherent imaging in the considered automotive scenario in terms of both probability of correct detection, evaluated considering several system parameters, as well as resolution capabilities, showcased by a dedicated experimental campaign wherein the collaboration between two vehicles enables the detection of the legs of a pedestrian close to a parked car. Moreover, as \textit{coherent} processing of several sensors' data requires very tight accuracy on clock synchronization and sensor's positioning -- referred to as \textit{phase synchronization} -- (such that to predict sensor-target distances up to a fraction of the carrier wavelength), we present a general three-step cooperative multistatic phase synchronization procedure, detailing the required information exchange among vehicles in the specific automotive radar context and assessing its feasibility and performance by hybrid Cram\'er-Rao bound.

Geosynchronous SARs (GEOSAR), studied since the 1970s, are missions providing huge coverage, subcontinental access within minutes, and near continuous observation capabilities. Such missions are suited for imaging and interferometric applications, aiming at large-scale monitoring of deformations, water vapor, and soil moisture. However, maintaining the orbit tube for interferometric requirements is quite demanding due to the perturbing forces at the GEO altitude. Effective control strategies are imperative to keep the baseline and the Doppler band and, ultimately, the orbit shape stable in time. This study investigates the feasibility of the near-zero inclination geostationary SAR concept, considering both interferometric requirements and compliance with the International Telecommunication Union (ITU) Regulation. The orbit maintenance problem is addressed, emphasizing the potential for persistent Earth observation in SAR interferometry applications.

This article discusses the effect of multipath in automotive radar imaging under different sensor configurations. The study is motivated by the fact that radar technologies are becoming indispensable in the automotive scenario. Many applications such as collision avoidance systems, assisted parking, and driving assistance systems take advantage of radar technologies to accomplish their task. However, one of the main concerns about automotive radars is the possibility of detecting false targets due to multiple signal reflections. In this article, we show how different sensor layouts experience multipath differently. In particular, we demonstrate that with multiple-input multiple-output (MIMO) radars, what really matters is the physical positions of the transmitting and receiving antennas. The monostatic/bistatic equivalent configurations cannot be used to design a system and to simulate an acquisition in the presence of a multipath. We also demonstrate how vehicle-based MIMO-synthetic aperture radar (MIMO-SAR) imaging can generate a bi-dimensional aperture which significantly reduces multipath effects in the focused image, avoiding the detection of false targets. All the theoretical analyses are supported by several simulations where different sensor layouts are tested, and the capability of MIMO-SAR to reject multipath is validated.

This paper analyzes the concept of multipath in automotive radar imaging, particularly in MIMO-SAR imaging. In a typical automotive environment, the radar signal may experience multiple reflections, which leads to the presence of ghost targets in the focused image. These targets may trigger undesired maneuvers from an advanced driving assistance system (ADAS), resulting in possible accidents. In this paper, we show how the position and brightness of the ghost targets are inherently related to the radar's physical layout, including the number of transmitting and receiving elements and their positions. Accordingly, we strongly recommend avoiding the usage of any monostatic or bistatic equivalence in simulation software since they will result in entirely erroneous results. We also show how MIMO-SAR, if implemented with the MIMO aperture orthogonal to the SAR aperture, is intrinsically robust to double bounces resulting in the suppression of ghost targets due to this effect. A set of simulations representing typical automotive scenarios support the theoretical analysis.

Detecting and monitoring changes in open-pit mines is crucial for efficient mining operations. Indeed, these changes comprise a broad spectrum of activities that can often lead to significant environmental impacts such as surface damage, air pollution, soil erosion, and ecosystem degradation. Conventional optical sensors face limitations due to cloud cover, hindering accurate observation of the mining area. To overcome this challenge, synthetic aperture radar (SAR) images have emerged as a powerful solution, due to their unique ability to penetrate clouds and provide a clear view of the ground. The open-pit mine change detection task presents significant challenges, justifying the need for a model trained for this specific task. First, different mining areas frequently include various features, resulting in a diverse range of land cover types within a single scene. This heterogeneity complicates the detection and distinction of changes within open-pit mines. Second, pseudo changes, e.g., equipment movements or humidity fluctuations, which show statistically reliable reflectivity changes, lead to false positives, as they do not directly correspond to the actual changes of interest, i.e., blasting, collapsing, or waste pile operations. In this paper, to the best of our knowledge, we present the first deep learning model in the literature that can accurately detect changes within open-pit mines using SAR images (TerraSAR-X). We showcase the fundamental role of data augmentations and a coherence layer as a critical component in enhancing the model’s performance, which initially relied solely on amplitude information. In addition, we demonstrate how, in the presence of a few labels, a pseudo-labeling pipeline can improve the model robustness, without degrading the performance by introducing misclassification points related to pseudo changes. The F1-Score results show that our deep learning approach is a reliable and effective method for SAR change detection in the open-pit mining sector.

With the advent of self-driving vehicles, autonomous driving systems will have to rely on a vast number of heterogeneous sensors to perform dynamic perception of the surrounding environment. Synthetic Aperture Radar (SAR) systems increase the resolution of conventional mass-market radars by exploiting the vehicle’s ego-motion, requiring very accurate knowledge of the trajectory, usually not compatible with automotive-grade navigation systems. In this setting, radar data are typically used to refine the navigation-based trajectory estimation with so-called autofocus algorithms. Although widely used in remote sensing applications, where the timeliness of the imaging is not an issue, autofocus in automotive scenarios calls for simple yet effective processing options to enable real-time environment imaging. This paper aims at providing a comprehensive theoretical and experimental analysis of the autofocus requirements in typical automotive scenarios. We analytically derive the effects of navigation-induced trajectory estimation errors on SAR imaging, in terms of defocusing and wrong targets’ localization. Then, we propose a motion estimation and compensation workflow tailored to automotive applications, leveraging a set of stationary Ground Control Points (GCPs) in the low-resolution radar images (before SAR focusing). We theoretically discuss the impact of the GCPs position and focusing height on SAR imaging, highlighting common pitfalls and possible countermeasures. Finally, we show the effectiveness of the proposed technique employing experimental data gathered during open road campaign by a 77 GHz multiple-input multiple-output radar mounted in a forward-looking configuration.

This paper proposes a method for efficient and accurate removal of grating lobes in automotive Synthetic Aperture Radar (SAR) images. Grating lobes can indeed be mistaken as real targets, inducing in this way false alarms in the target detection procedure. Grating lobes are present whenever SAR focusing is performed using data acquired on a non-continuous basis. This kind of acquisition is typical in the automotive scenario, where regulations do not allow for a continuous operation of the radar. Radar pulses are thus transmitted and received in bursts, leading to a spectrum of the signal containing gaps. We start by deriving a suitable reference frame in which SAR images are focused. It will be shown that working in this coordinate system is particularly convenient since it allows for a signal spectrum that is space-invariant and with spectral gaps described by a simple one-dimensional function. After an inter-burst calibration step, we exploit these spectral characteristics of the signal by implementing a compressive sensing algorithm aimed at removing grating lobes. The proposed approach is validated using real data acquired by an eight-channel automotive radar operating in burst mode at 77 GHz. Results demonstrate the practical possibility to process a synthetic aperture length as long as up to 2 m reaching in this way extremely fine angular resolutions.

Automotive Synthetic Aperture Radar (SAR) is a promising technology for autonomous driving, where reliable perception of the environment is needed. Though, SAR focusing needs precise vehicle’s trajectory knowledge, not compatible with automotive-grade navigation systems. Current autofocus algorithms refine navigation-based trajectory with radar data but do not exploit vehicle’s dynamic in the residual motion estimation. This paper investigates the injection of a-priori knowledge into residual motion estimation to achieve improved and physically consistent SAR imaging. An autoregressive model of the residual velocities and Bayesian tracking via Kalman Filter are proposed and deeply studied upon application on real data acquired in an open road campaign. A new metric is introduced to quantitatively compare the outcomes: the variance of Hough lines angular coefficients. Experimental results confirm that the metric is informative, and the presence of memory in the residual motion estimation is effective in better estimating residual velocity and, consequently, improved SAR imaging.

This paper tackles two open problems concerning automotive Synthetic Aperture Radar (SAR): whether it is possible to develop a fast and efficient focusing routine and assess the effect of multipath on MIMO-SAR images. First, we present a processing technique that enables real-time imaging of the scene under observation. The idea is that the SAR image is already present in the Range-Angle-Velocity (RAV) data cube and must be extracted with a simple 3D interpolation. Unlike standard Doppler beam sharpening techniques, this processor is accurate since it handles range migration and phase curvature. For what concerns the multipath, instead, we introduce a simple yet effective approach to mitigate this issue, which significantly improves the robustness of SAR systems concerning multipath. To showcase the capabilities and potential of our proposed method, we conducted a comprehensive set of experiments utilizing simulated data. The results were not only able to demonstrate the enhancement of robustness of the SAR system in dealing with multipath but also highlighted the ability of our approach to deliver highly accurate and high-quality real-time imaging of the scene being surveyed.

The paper proposes a flexible and efficient wavenumber domain processing scheme suited for close formations of low earth orbiting (LEO) synthetic aperture radar (SAR) sensors hosted on micro-satellites or CubeSats. Such systems aim to generate a high-resolution image by combining data acquired by each sensor with a low pulse repetition frequency (PRF). This is usually performed by first merging the different channels in the wavenumber domain, followed by bulk focusing. In this paper, we reverse this paradigm by first upsampling and focusing each acquisition and then combining the focused images to form a high-resolution, unambiguous image. Such a procedure is suited to estimate and mitigate artifacts generated by incorrect positioning of the sensors. An efficient wave–number method is proposed to focus data by adequately coping with the orbit curvature. Two implementations are provided with different quality/efficiency. The image quality in phase preservation, resolution, sidelobes, and ambiguities suppression is evaluated by simulating both point and distributed scatterers. Finally, a demonstration of the capability to compensate for ambiguities due to a small across-track baseline between sensors is provided by simulating a realistic X-band multi-sensor acquisition starting from a stack of COSMO-SkyMed images.

Automotive Synthetic Aperture Radar (SAR) imaging is becoming increasingly relevant in the automotive industry thanks to its day and night, all-weather, and high-resolution imaging capabilities. The latter, in particular, is achieved by jointly processing several radar pulses. Since the vehicle where the radar is mounted is moving, each pulse illuminates the scene from a slightly different spatial location, generating bandwidth and, in turn, resolution. To achieve extremely fine resolutions, however, it is mandatory to use very long synthetic apertures; in other words, it is necessary to process all the radar pulses acquired on a very long portion of the vehicle's trajectory. Nevertheless, a radar that is fully compliant with the regulations must avoid a continuous transmission of pulses. The device transmits a burst of pulses, and then it must remain silent for a certain period. While this condition is not a problem for standard Multiple- Input Multiple-Output (MIMO) radar imaging, it is a big issue for synthetic aperture imaging. In the latter case, correct imaging can happen only when all the pulses sample uniformly the synthetic aperture, but this is impossible when pulses are transmitted on a non-uniform basis. Therefore, if we want to achieve extremely fine resolutions, we have to use very long apertures that are non-uniformly sampled. The uneven sampling generates gaps in the image spectrum with the consequent generation of side lobes. These side lobes in SAR images can eventually be mistaken for real targets triggering in this way unwanted maneuvers by an Advanced Driving Assistance System (ADAS). This work presents a novel method to eliminate side-lobes from high-resolution SAR images: the procedure starts by defining an ad-hoc reference system in which the spectral components of the image are independent of the position of the targets in the scene. This reference system also allows the description of the spectral gaps by a simple mono-dimensional function. After that, we exploit a well-known compressive sensing algorithm called CLEAN to remove side lobes. The proposed approach is validated using real data from an 8-channel automotive Radar operating in burst mode at 77 GHz. Results demonstrate the practical possibility of processing a synthetic aperture length as long as 2 meters, reaching outstanding angular resolutions.

In this paper, we discuss the possibility of generating high-resolution mapping of urban (or extra-urban) environments by the application of synthetic aperture radar (SAR) processing concepts to the data collected by mm-wave automotive radars installed on-board commercial vehicles. The study is motivated by the fact that radar sensors are becoming an indispensable component of the equipment of modern vehicles, being characterized by low cost, good performance, and affordable processing; therefore, in the future, nearly every single vehicle could be potentially equipped with radar devices capable of high-resolution imaging, enabled by application of SAR processing methodologies. Throughout this paper, we aim to discuss the role of SAR imaging in the automotive context under a theoretical and experimental perspective. First, we present the resulting benefits in terms of angular resolution and signal-to-noise ratio. Then, we discuss relevant technological aspects, such as suppression of angular ambiguities, fine estimation of platform motion, and SAR processing architectures, and we present a preliminary evaluation of the required computational costs. Finally, we will present a number of experimental results based on open road campaign data acquired using an 8-channel MIMO radar at 77 GHz, considering the cases of side-looking SAR, forward SAR, and SAR imaging of moving targets.

The principal element of interest concerning the use of Synthetic Aperture Radar (SAR) technology in the automotive scenario is the possibility to synthesize an arbitrarily long array by exploiting the natural motion of the ego-vehicle, and therefore achieve much finer spatial resolution and improved detection capabilities without increasing hardware requirements in terms of number of physical antennas. In this paper, we discuss the application of SAR imaging in the automotive context under a theoretical and experimental perspective. Experimental results are shown based on open road campaign data acquired using an 8-channel Radar at 77 GHz, considering the cases of side-looking SAR, forward SAR, and SAR imaging of moving targets. Results corroborate the idea that SAR imaging could be successfully and systematically used in the near future for high-resolution mapping of the urban environment.

Automotive synthetic aperture radar (SAR) systems are rapidly emerging as a candidate technological solution to enable a high-resolution environment mapping for autonomous driving. Compared to lidars and cameras, automotive-legacy radars can work in any weather condition and without an external source of illumination, but are limited in either range or angular resolution. SARs offer a relevant increase in angular resolution, provided that the ego-motion of the radar platform is known along the synthetic aperture. In this paper, we present the results of an experimental campaign aimed at assessing the potential of a multi-beam SAR imaging in an urban scenario, composed of various targets (buildings, cars, pedestrian, etc.), employing a 77 GHz multiple-input multiple-output (MIMO) radar platform based on a mass-market available automotive-grade technology. The results highlight a centimeter-level accuracy of the SAR images in realistic driving conditions, showing the possibility to use a multi-angle focusing approach to detect and discriminate between different targets based on their angular scattering response.

In this work we describe a methodology to process the data acquired by a conventional mm-wave automotive Radar to form an image of static targets ahead of the ego-vehicle at a spatial resolution several times finer than allowed by the physical size of the Radar array. High-resolution imaging is achieved by applying a processing scheme typical of Synthetic Aperture Radars (SAR), that is by exploiting the forward motion of the ego-vehicle to synthesize an aperture as long as several centimeters. Contrary to many works in literature, where SAR imaging is achieved at the expense of a significant increase of computational costs, we propose here a Quick and Dirty (Q&D) approach that can be implemented at computational costs comparable with those of standard automotive Radar imaging. The effectiveness of this approach is demonstrated on the basis of real data collected during an open road acquisition campaign, by comparing Q&D SAR images with accurate SAR images produced by Time Domain Back Projection. Finally, it is shown that the algorithm can easily be tuned to produce SAR imaging of moving taraets.

Synthetic Aperture Radar (SAR) imaging is starting to play an essential role in the automotive industry. Its day and night sensing capability, fine resolution, and high flexibility are key aspects making SAR a very compelling instrument in this field. This paper describes and compares three algorithms used to combine low-resolution images acquired by a Multiple-Input Multiple-Output (MIMO) automotive radar to form an SAR image of the environment. The first is the well-known Fast Factorized Back-Projection (FFBP), which focuses the image in different stages. The second one will be called 3D2D, and it is a simple 3D interpolation used to extract the SAR image from the Range-Angle-Velocity (RAV) data cube. The third will be called Quick&Dirty (Q&D), and it is a fast alternative to the 3D2D scheme that exploits the same intuition. A rigorous mathematical description of each algorithm is derived, and their limits are addressed. We then provide simulated results assessing different interpolation kernels, proving which one performs better. A rough estimation of the number of operations proves that both algorithms can be deployed using a real-time implementation. Finally, we will present some experimental results based on open road campaign data acquired using an eight-channel MIMO radar at 77 GHz, considering the case of a forward-looking geometry.

This paper deals with the analysis, estimation, and compensation of trajectory errors in automotive-based Synthetic Aperture Radar (SAR) systems. First of all, we define the geometry of the acquisition and the model of the received signal. We then proceed by analytically evaluating the effect of an error in the vehicle's trajectory. Based on the derived model, we introduce a procedure capable of estimating and compensating constant velocity motion errors leading to a well-focused and well-localized SAR image. The procedure is validated using real data gathered by a 77 GHz automotive SAR with MIMO capabilities.