Search

International audience; This paper explains CentraleSupélec team's approach for the challenge of human activity classification with radar [1]. Our methodology is based on a convolutional network classifier processing the range-time and spectrogram representations. The preprocessing and the neural network architecture presented in this paper allowed us to reach 99% accuracy on challenge's test data of the challenge. A more advanced preprocessing using wavelets is also briefly discussed. Index Terms-Human activity classification, deep convolutional neural networks, spectrogram, Doppler-range diagram, rangetime diagram.

As opposed to metasurfaces, metagratings represent themselves sparse arrangements of scatterers. Established rigorous analytical models allow metagratings to overcome performance of metasurfaces in beam steering applications while handling less degrees of freedom. In this work we deal with reflective metagratings that have only as few as one degree of freedom (represented by a reactively loaded thin wire) per each propagating diffraction order. We present a detailed design procedure and fabrication of three experimental samples capable of establishing prescribed diffraction patterns. The samples are experimentally studied in an anechoic chamber dedicated to radar-cross-section bistatic measurements and results are compared with three-dimension full wave numerical simulations. We identify and analyze factors affecting operating frequency range of metagratings, suggest a strategy to increase the bandwidth., Comment: 8 pages, 5 figures

International audience; Recently, metamaterial-inspired diffraction gratings (or metagratings) have demonstrated unprecedented efficiency in wavefront manipulation by means of relatively simple structures. Conventional one-dimensional (1D) gratings have a profile modulation in one direction and a translation symmetry in the other. In 1D metagratings, the translation invariant direction is engineered at a subwavelength scale, which allows one to accurately control polarization line currents and, consequently, the scattering pattern. In this paper, we present the design and experimental results of a reconfigurable metagrating operating at 14 GHz. The unit cell of the metagrating incorporates a varactor diode as a tunable element and an additional passive inductive grid is used to effectively decrease the minimum capacitance of the varactor. By varying the applied bias voltage, one is able to perform beam steering as well as generate multiple beams with a single feeding horn antenna.

International audience; Covariance matrix estimation is a ubiquitous problem in signal processing. In most modern signal processing applications, data are generally modeled by non-Gaussian distributions with covariance matrices exhibiting a particular structure. Taking into account this structure and the non-Gaussian behavior improve drastically the estimation accuracy. In this paper, we consider the estimation of structured scatter matrix for complex elliptically distributed observations, where the assumed model can differ from the actual distribution of the observations. Specifically, we tackle this problem, in a mismatched framework, by proposing a novel estimator, named StructurEd ScAtter Matrix Estimator (SESAME), which is based on a two-step estimation procedure. We conduct theoretical analysis on the unbiasedness and the asymptotic efficiency and Gaussianity of SESAME. In addition, we derive a recur-sive estimation procedure that iteratively applies the SESAME method, called Recursive-SESAME (R-SESAME), reaching with improved performance at lower sample support the (Mismatched) Cramér-Rao Bound. Furthermore, we show that some special cases of the proposed method allow to retrieve preexisting methods. Finally, numerical results corroborate the theoretical analysis and assess the usefulness of the proposed algorithms.

Given a double-bounce model, an analytical solution to retrieve the complex effective permittivities from a dihedral corner made of two materials is proposed using backscattered electromagnetic fields in dual polarization. Domains of validity for this solution are provided, and a simpler method for a homogeneous corner is presented. The general process is successfully tested on simulated data performed using geometric optics applied to a dihedral arrangement from 0.3 to 1 GHz. It allows in the monostatic case to get the materials in urban or forest environments.

The Double Brewster angle effect (DBE) is an extension of the Brewster angle to double reflection on two orthogonal dielectric surfaces. It results from the combination of two pseudo-Brewster angles occurring in complementary incidence angles domains. It can be observed for a large range of incidence angles provided that double bounces mechanism is present. As a consequence of this effect, we show that the reflection coefficient at VV polarization can be at least 10 dB lower than the reflection coefficient at HH polarization over a wide range of incidence angle – typically from 20 to 70 ∘ . It is experimentally demonstrated using a Synthetic Aperture Radar (SAR) image that this effect can be seen on buildings and forests. For large buildings, the difference can reach more than 20 dB.

Remote sensing data from Synthetic Aperture Radar (SAR) sensors offer a unique opportunity to record, to analyze, and to predict the evolution of the Earth. In the last decade, numerous satellite remote sensing missions have been launched (Sentinel-1, UAVSAR, TerraSAR X, etc.). This resulted in a dramatic improvement in the Earth image acquisition capability and accessibility. The growing number of observation systems allows now to build high temporal/spatial-resolution Earth surface images data-sets. This new scenario significantly raises the interest in time-series processing to monitor changes occurring over large areas. However, developing new algorithms to process such a huge volume of data represents a current challenge. In this context, the present thesis aims at developing methodologies for change detection in high-resolution SAR image time series.These series raise two notable challenges that have to be overcome:On the one hand, standard statistical methods rely on multivariate data to infer a result which is often superior to a monovariate approach. Such multivariate data is however not always available when it concerns SAR images. To tackle this issue, new methodologies based on wavelet decomposition theory have been developed to fetch information based on the physical behavior of the scatterers present in the scene.On the other hand, the improvement in resolution obtained from the latest generation of sensors comes with an increased heterogeneity of the data obtained. For this setup, the standard Gaussian assumption used to develop classic change detection methodologies is no longer valid. As a consequence, new robust methodologies have been developed considering the family of elliptical distributions which have been shown to better fit the observed data.The association of both aspects has shown promising results in change detection applications., La télédétection par Radar à Synthèse d’Ouverture (RSO) offre une opportunité unique d’enregistrer, d’analyser et de prédire l’évolution de la surface de la Terre. La dernière décennie a permis l’avènement de nombreuses missions spatiales équipées de capteurs RSO (Sentinel-1, UAVSAR, TerraSAR X, etc.), ce qui a engendré une rapide amélioration des capacités d’acquisition d’images de la surface de la Terre. Le nombre croissant d’observations permet maintenant de construire des bases de données caractérisant l’évolution temporelle d’images, augmentant considérablement l’intérêt de l’analyse de séries temporelles pour caractériser des changements qui ont lieu à une échelle globale. Cependant, le développement de nouveaux algorithmes pour traiter ces données très volumineuses est un défi qui reste à relever. Dans ce contexte, l’objectif de cette thèse consiste ainsi à proposer et à développer des méthodologies relatives à la détection de changements dans les séries d’images ROS à très haute résolution spatiale.Le traitement de ces séries pose deux problèmes notables. En premier lieu, les méthodes d’analyse statistique performantes se basent souvent sur des données multivariées caractérisant, dans le cas des images RSO, une diversité polarimétrique, interférométrique, par exemple. Lorsque cette diversité n’est pas disponible et que les images RSO sont monocanal, de nouvelles méthodologies basées sur la décomposition en ondelettes ont été développées. Celles-ci permettent d’ajouter une diversité supplémentaire spectrale et angulaire représentant le comportement physique de rétrodiffusion des diffuseurs présents la scène de l’image. Dans un second temps, l’amélioration de la résolution spatiale sur les dernières générations de capteurs engendre une augmentation de l’hétérogénéité des données obtenues. Dans ce cas, l’hypothèse gaussienne, traditionnellement considérée pour développer les méthodologies standards de détection de changements, n’est plus valide. En conséquence, des méthodologies d’estimation robuste basée sur la famille des distributions elliptiques, mieux adaptée aux données, ont été développées.L’association de ces deux aspects a montré des résultats prometteurs pour la détection de changements.Le traitement de ces séries pose deux problèmes notables. En premier lieu, les méthodes d’analyse statistique performantes se basent souvent sur des données multivariées caractérisant, dans le cas des images RSO, une diversité polarimétrique ou interférométrique, par exemple. Lorsque cette diversité n’est pas disponible et que les images RSO sont monocanal, de nouvelles méthodologies basées sur la décomposition en ondelettes ont été développées. Celles-ci permettent d’ajouter une diversité spectrale et angulaire supplémentaire représentant le comportement physique de rétrodiffusion des diffuseurs présents la scène de l’image. Dans un second temps, l’amélioration de la résolution spatiale sur les dernières générations de capteurs engendre une augmentation de l’hétérogénéité des données obtenues. Dans ce cas, l’hypothèse gaussienne, traditionnellement considérée pour développer les méthodologies standards de détection de changements, n’est plus valide. En conséquence, des méthodologies d’estimation robuste basée sur la famille des distributions elliptiques, mieux adaptée aux données, ont été développées.L’association de ces deux aspects a montré des résultats prometteurs pour la détection de changements.

Although a rigorous theoretical ground on metasurfaces has been established in the recent years on the basis of the equivalence principle, the majority of metasurfaces for converting a propagating wave into a surface wave are developed in accordance with the so-called generalized Snell's law being a simple heuristic rule for performing wave transformations. Recently, for the first time, Tcvetkova et al. [Phys. Rev. B 97, 115447 (2018)] have rigorously studied this problem by means of a reflecting anisotropic metasurface, which is, unfortunately, difficult to realize, and no experimental results are available. In this paper, we propose an alternative practical design of a metasurface-based converter by separating the incident plane wave and the surface wave in different half-spaces. It allows one to preserve the polarization of the incident wave and substitute the anisotropic metasurface by an omega-bianisotropic one. The problem is approached from two sides: By directly solving the corresponding boundary problem and by considering the ``time-reversed'' scenario when a surface wave is converted into a nonuniform plane wave. We develop a practical three-layer metasurface based on a conventional printed circuit board technology to mimic the omega-bianisotropic response. The metasurface incorporates metallic walls to avoid coupling between adjacent unit cells and accelerate the design procedure. The design is validated with full-wave three-dimensional numerical simulations and demonstrates high conversion efficiency., 10 pages, 6 figures, 1 table

International audience; This paper addresses robust estimation of structured shape (normalized covariance) matrices. Shape matrices most often own a particular structure depending on the application of interest and taking this structure into account improves estimation accuracy. In the framework of robust estimation, we introduce a recursive robust shape matrix estimation technique based on Tyler's M-estimate for convexly structured shape matrices. We prove that the proposed estimator is consistent, asymptotically efficient and Gaussian distributed and we notice that it reaches its asymptotic regime faster as the number of recursions increases. Finally, in the particular wide spreaded case of Hermitian persymmetric structure, we study the convergence of the recursions of the proposed algorithm.

International audience; Passive radar systems generally receive direct path, ground echoes and potential target echoes. As the first two are much more powerful than the target signals, they have to be rejected. Numerous solutions are proposed in the literature. The algorithm introduced by [1] has become a reference. Due to the airborne passive clutter spread, this algorithm faces limits. In this article, we derive two solutions in order to be able to reject a large amount of clutter and therefore to reduce detrimental secondary sidelobes. Experimental results prove that the proposed method performs only 4 dB worse than in a ground configuration.

International audience; Time Modulated Array offers a simple way to synthesize radiation patterns with low sidelobe levels, in time-average, using an optimized strategy of switching ON and OFF the elements of the array. In this work we show that the complexity of the optimization strategy can be reduced by first shrinking the search space to a subset, or database, of useful solutions over which the optimization will be launched. This reduction is obtained by applying the constraints in a progressive and hierarchical order. In this way, this problem of exponential growth is reduced to a polynomial growth, with respect to the size of the antenna array. We present the case of designing the time modulated array for radar applications where constant directivity and ability of intereference rejection is needed.

The purpose of this paper is to present a study of non-Gaussian detectors for the detection of small, slow moving targets in clutter. These detectors belong to the family of Adaptive Normalized Matched Filter. The noise-clutter covariance matrix will be computed by the classic fixed point estimator [1, 2, 3] or with an iterative estimator based on the multi-segment Burg algorithm [4]. We will also propose to add a data selection algorithm based on order statistics in order to improve the estimation of this covariance matrix when targets are in the clutter alone reference cells

International audience; Radio beacons, at around 115 MHz (VHF, VOR), are used as opportunity donors in a passive radar configuration. Full-wave simulation of three categories (according to size) of aircrafts shows that their bistatic RCS are statistically comparable without any additional information but perform differently in time (while flying). This discrepancy can be exploited to recognize the aircrafts among the categories. Measurements confirm these ideas and indicate that it may be possible to differentiate the aircrafts among the same category. Encouraging initial results were obtained using Convolutional Neural Networks to classify airplane categories combining simulated bistatic RCS results and real trajectories (collected from ADS-B data).

Ce papier porte sur l'estimation robuste de matrices de forme (covariance normalisée) structurées. Dans un contexte d'estimation robuste, nous introduisons un estimateur robuste et récursif de matrice de forme basé sur l'estimateur de Tyler. Nous prouvons que cet estimateur est consistant, asymptotiquement efficace et gaussien. Par ailleurs, nous constatons qu'il atteint son régime asymptotique d'autant plus rapidement que le nombre d'itérations augmente. Finalement, nous étudions la convergence des récursions pour la structure persymétrique hermitienne., This paper addresses robust estimation of structured shape (normalized covariance) matrices. In the framework of robust estimation, we introduce a recursive robust shape matrix estimation technique based on Tyler's M-estimate for convexly structured shape matrices. We prove that the proposed estimator is consistent, asymptotically efficient and Gaussian distributed and we notice that it reaches its asymptotic regime faster as the number of recursions increases. Finally, in the case of Hermitian persymmetric structure, we study the convergence of the recursions of the proposed algorithm.

This paper is interested in phase code design, and especially in their robustness to the eclipsing phenomenon. In a radar background, the term "eclipsing" describes a partial loss of data at the reception side. It implies an increase of pulse compression sidelobes, and thus a degradation of detection performance. In this regard, this paper describes an optimisation procedure which aims to provide waveforms (phase codes) that are quite robust to the eclipsing phenomenon. Some simulations are provided to highlight the usefulness of the procedure., Dans ce papier, on s'intéresse aux codes de phase, et plus particulièrement à leur robustesse aux phénomènes d'éclipse. Dans le contexte radar, une éclipse se traduit par une perte partielle d'informations, due à la réception d'un signal pendant la phase d'émission. Cela implique notamment une élévation du niveau des lobes secondaires après compression d'impulsion, et par conséquent une dégradation des performances de détection. En ce sens, ce papier propose une procédure d'optimisation permettant de définir des codes de phase robustes aux phénomènes d'éclipse. Quelques simulations soulignent l'intérêt de la méthodologie.

This paper considers the problem of detecting changes in multivariate Synthetic Aperture Radar image time series. Classical methodologies based on covariance matrix analysis are usually built upon the Gaussian assumption, as well as an unstructured signal model. Both of these hypotheses may be inaccurate for high-dimension/resolution images, where the noise can be heterogeneous (non-Gaussian) and where all channels are not always informative (low-rank structure). In this paper, we tackle these two issues by proposing a new detector assuming a robust low-rank model. Analysis of the proposed method on a UAVSAR dataset shows promising results., Ce papier considère la détection de changements dans une série temporelle d'images multivariées obtenues par radarà synthèse d'ouverture. Dans ce cadre les méthodologies classiques modélisent les données par une distribution gaussienneà moyenne nulle et une cova-riance non structurée. Ces deux hypothèses montrent leur limite lorsque les données sontà très haute résolution ou que la dimension de celles-ci est grande. Pour de telles images, les pixels peuventêtre hétérogènes (non gaussiens) et peuvent avoir une structure de rang faible. Ainsi, dans ce papier nous proposons de dériver un nouveau détecteur dans un cadre robuste en supposant une structure de rang faible. Les résultats de l'approche proposée sur un jeu de données UAVSAR illustrent l'intérêt de la méthode.

This article illustrates the benefit of crosspolarization, both in intensity and in phase, to discriminate urban areas and vegetation when there is an ambiguity while using the Pauli decomposition. This ambiguity occurs when buildings are rotated with respect to the sensor trajectory. In this case, we show that because the dissymetry is not of the same nature (deterministic vs quasi-random), it is then possible to discriminate these two environments.

International audience; This paper considers the problem of detecting changes in mul-tivariate Synthetic Aperture Radar image time series. Classical methodologies based on covariance matrix analysis are usually built upon the Gaussian assumption, as well as an unstructured signal model. Both of these hypotheses may be inaccurate for high-dimension/resolution images, where the noise can be heterogeneous (non-Gaussian) and where all channels are not always informative (low-rank structure). In this paper, we tackle these two issues by proposing a new detector assuming a robust low-rank model. Analysis of the proposed method on a UAVSAR dataset shows promising results .

Wild fire is an inflexible threat for the environment equilibrium and human lives, consequently it is vital to study its behaviour for the collective interest and scientific knowledge. Satellite and airbone SAR represents an interesting alternative to optical sensors when the conditions are not favorable. For instance, with presence of clouds, heavy rain or smoke the efficiency of optical sensors decrease drastically for imagery purpose. Since SAR sensors overcome these issues, it is interesting to evaluate their capability to detect fire burn-scars as well as active fires. The NASA Jet Propulsion Lab provide airbone multitemporal L-band full polarimetric data for several areas in USA and South America, in particular, some data are available in California where countrysides are know to be affected by regular significant fires. Depending on the severity of the fire and the vegetation, different physical mechanisms might be affected. The aim of this study is to evaluate the benefit of Pauli basis decomposition in the frame of Change Detection algorithm for burn scar detection.

International audience

International audience; Confusion between the polarimetric radar signatures of rotated buildings and vegetation has been widely studied to compensate for the effect of rotation. We propose in this study to have a new look on this problem. To do this, we have selected different areas in the San Francisco bay and in New Orleans. These selected zones present different orientations with respect to the sensor illuminating the scene, RADARSAT-2 (C-band) for San Francisco or UAVSAR (L-band) for New Orleans. The variations with the orientation angle of the VV/HH and HV/HH responses collected over these areas are almost identical whatever the sensor. These two quantities first grow and rapidly reach a plateau, at a level which is similar to the responses of forests we collected throughout literature. Actually, it seems that, when the rotation angles grows, the polarimetric radar responses of urban areas tend to limit values which are similar to the radar responses of forests.

This thesis deals with coherent MIMO radars. MIMO stands for Multiple-Input Multiple-Output, meaning that several transmitters and several receivers are used, closely-spaced in a coherent MIMO radar. Each transmitter has its own signal, providing waveform diversity. This thesis aims for defining a way to generate a set of sequences, specific for this radar, while satisfying practical constraints. It may help to determine the potential contribution of a MIMO radar. Only phase codes are concerned here, because they suffer less from the range/angle coupling effect.A simple framework is introduced, based on an optimisation problem.While literature often involves the orthogonality of the elementary signals (because of theoretical aspects), it is suggested to consider the orthogonality of signals from different directions of the surveillance space. Unfortunately, the obtained optimisation problem is non-convex and has a lot of variables. A benchmark on a simpler problem notifies us that gradient-based algorithms are surprisingly efficient.An optimisation of the correlation function corresponds to a processing with matched filters. However, in practice, mismatched filters are usually employed. A joint optimisation problem is suggested accordingly, in order to generate simultaneously a set of sequences (e.g. MIMO radar signals) and their associated mismatched filters. Obtained results are quite promising : as expected, a joint optimisation seems to perform better than a cyclic one, usually employed.; Cette thèse se focalise sur le concept du radar MIMO co-localisé. L'acronyme MIMO -- pour Multiple-Input Multiple-Output -- indique l'utilisation de plusieurs émetteurs et de plusieurs récepteurs, tandis que le terme co-localisé signifie que ces éléments sont étroitement espacés. Chaque émetteur envoie une forme d'onde qui lui est propre : un radar MIMO émet donc simultanément un ensemble de signaux.Cette thèse a ainsi pour but d'établir une méthodologie permettant de générer cet ensemble de signaux, tout en respectant certaines contraintes opérationnelles. Cela nous permettra de déterminer les apports éventuels de ce radar. Nous nous sommes intéressés en particulier aux codes de phase, pour des raisons de couplage (qu'on peut traduire ici par la capacité, lors du traitement, à distinguer la position angulaire d'une cible de sa distance).La méthodologie proposée se synthétise simplement en une modélisation sous la forme d'un problème d'optimisation. Contrairement à la littérature et à des précédents résultats théoriques, nous avons décidé d'évaluer l'orthogonalité des signaux émis par le radar en différentes directions, et non l'orthogonalité des signaux élémentaires. Ce problème, plus réaliste, est malheureusement non-convexe et à grande échelle : un benchmark sur différentes méthodes d'optimisation nous a permis de constater l'efficacité des algorithmes basées sur le gradient.Optimiser cette orthogonalité sous-entend l'utilisation de filtres adaptés. Cependant, en pratique, le traitement radar s'effectue à l'aide de filtres désadaptés. Nous suggérons ainsi un problème d'optimisation jointe, permettant de générer de manière simultanée un ensemble de formes d'onde (pour le radar MIMO, entre autres) et les filtres désadaptés associés. Des simulations ont permis de montrer l'efficacité de la méthode. Celle-ci est en particulier préférable aux algorithmes cycliques habituellement utilisés.

This thesis deals with coherent MIMO radars. MIMO stands for Multiple-Input Multiple-Output, meaning that several transmitters and several receivers are used, closely-spaced in a coherent MIMO radar. Each transmitter has its own signal, providing waveform diversity. This thesis aims for defining a way to generate a set of sequences, specific for this radar, while satisfying practical constraints. It may help to determine the potential contribution of a MIMO radar. Only phase codes are concerned here, because they suffer less from the range/angle coupling effect.A simple framework is introduced, based on an optimisation problem.While literature often involves the orthogonality of the elementary signals (because of theoretical aspects), it is suggested to consider the orthogonality of signals from different directions of the surveillance space. Unfortunately, the obtained optimisation problem is non-convex and has a lot of variables. A benchmark on a simpler problem notifies us that gradient-based algorithms are surprisingly efficient.An optimisation of the correlation function corresponds to a processing with matched filters. However, in practice, mismatched filters are usually employed. A joint optimisation problem is suggested accordingly, in order to generate simultaneously a set of sequences (e.g. MIMO radar signals) and their associated mismatched filters. Obtained results are quite promising : as expected, a joint optimisation seems to perform better than a cyclic one, usually employed., Cette thèse se focalise sur le concept du radar MIMO co-localisé. L'acronyme MIMO -- pour Multiple-Input Multiple-Output -- indique l'utilisation de plusieurs émetteurs et de plusieurs récepteurs, tandis que le terme co-localisé signifie que ces éléments sont étroitement espacés. Chaque émetteur envoie une forme d'onde qui lui est propre : un radar MIMO émet donc simultanément un ensemble de signaux.Cette thèse a ainsi pour but d'établir une méthodologie permettant de générer cet ensemble de signaux, tout en respectant certaines contraintes opérationnelles. Cela nous permettra de déterminer les apports éventuels de ce radar. Nous nous sommes intéressés en particulier aux codes de phase, pour des raisons de couplage (qu'on peut traduire ici par la capacité, lors du traitement, à distinguer la position angulaire d'une cible de sa distance).La méthodologie proposée se synthétise simplement en une modélisation sous la forme d'un problème d'optimisation. Contrairement à la littérature et à des précédents résultats théoriques, nous avons décidé d'évaluer l'orthogonalité des signaux émis par le radar en différentes directions, et non l'orthogonalité des signaux élémentaires. Ce problème, plus réaliste, est malheureusement non-convexe et à grande échelle : un benchmark sur différentes méthodes d'optimisation nous a permis de constater l'efficacité des algorithmes basées sur le gradient.Optimiser cette orthogonalité sous-entend l'utilisation de filtres adaptés. Cependant, en pratique, le traitement radar s'effectue à l'aide de filtres désadaptés. Nous suggérons ainsi un problème d'optimisation jointe, permettant de générer de manière simultanée un ensemble de formes d'onde (pour le radar MIMO, entre autres) et les filtres désadaptés associés. Des simulations ont permis de montrer l'efficacité de la méthode. Celle-ci est en particulier préférable aux algorithmes cycliques habituellement utilisés.

A growing problem in the remote sensing community concerns the estimation of change-points in a time series of Synthetic Aperture Radar (SAR) images. Although the methodologies of change-point estimation have already been investigated in the literature, there are, to the best of our knowledge, no study on the expected performance for the estimation of change-points in a Wishart distributed time series. This is mainly due to the fact that few results exist on change-point estimation performance in the mathematical literature: the classical central limit theorem does not apply and the classical Cramer-Rao Bound does not exist due to the discrete nature of the parameters. To fill this gap, this paper proposes to use a lower-bound on the Mean Square Error (MSE) with fewer regularity conditions. To this end, recent works on hybrid Cramer-Rao/Weiss-Weinstein bound have been adapted to the specific SAR problematic of interest. Since estimation strategies usually rely on a set of parameters which have to be set by the user, we show how the proposed lower bound allows performing an appropriate tuning. Moreover, the proposed bound is computationally efficient which enables an extensive analysis without a high computational cost.

Recently, metamaterials-inspired diffraction gratings (or metagratings) have demonstrated unprecedented efficiency in wavefront manipulation by means of relatively simple structures. Conventional one-dimensional (1D) gratings have a profile modulation in one direction and a translation symmetry in the other. In 1D metagratings, the translation invariant direction is engineered at a subwavelength scale, which allows one to accurately control polarization line currents and, consequently, the scattering pattern. In bright contrast to metasurfaces, metagratings cannot be described by means of surface impedances (or local reflection and transmission coefficients). In this paper, we present a simulation-based design approach to construct metagratings in the "unit cell by unit cell" manner. It represents an analog of the local periodic approximation that has been used to design space-modulated metasurfaces and allows one to overcome the limitations of straightforward numerical optimization and semianalytical procedures that have been used to date to design metagratings. Electric and magnetic metagrating structures responding to, respectively, transverse electric and transverse magnetic incident plane waves are presented to validate the proposed design approach., Comment: 11 pages, 6 figures

Given a target prior information, our goal is to propose a method for automatically separating targets of interests from the background in hyperspectral imagery. More precisely, we regard the given hyperspectral image (HSI) as being made up of the sum of low-rank background HSI and a sparse target HSI that contains the targets based on a pre-learned target dictionary constructed from some online spectral libraries. Based on the proposed method, two strategies are briefly outlined and evaluated to realize the target detection on both synthetic and real experiments., Some wrong information present in the published IEEE version are corrected here (marked in red)

International audience; Detection and localization in urban environments is a very recent radar problem. In this article, we investigate the possibility of detecting and locating targets in NLOS areas with a single portable radar by exploiting multipath returns. We propose two algorithms which handle the information provided by multipath returns in different ways to detect and estimate the NLOS target position. We also present an original method to select the number of paths to take into account in the algorithms in order to maximize detection probabilities. Numerical results show good efficiency of the proposed algorithms for problems of both detection and localization. We show that applying these algorithms improves detection performance compared to a classic matched filter in a typical urban scenario. Experimental results on a real data set allow us to validate our multipath model in urban environments, and in particular to show that it is possible to retrieve the target location even with rough knowledge of the scene geometry.

For a monostatic radar, the dihedral arrangementexhibits a strong response compare tobare surface. This phenomenon can be seen in radarimages of urban areas or forestry as building andstreets and trunks and ground are forming such structures.This signature can be modelled with the doublebounce equation as a cascade of Fresnel reflectioncoefficients. These coefficients depend on the permittivitiesof the structures. Finding these permittivitiesfrom the backscattered fields may allowed to identifythe materials involved in the dihedral. Thereforethiscan be applied for default detection, moisture retrievaland classification topics for instance.By modelling the double bounce scattering with Fresnelcoefficients, three methods have been developedto retrieve these permittivities. The first one uses asingle incidence angle and the measurements of copolarisedfields. The last two methods need the measurementof the copolarised fields ratio or of the horizontalelectric field alone at two different incidenceangles. All these methods have been validated, firstnumerically using FEKO RL-GO method, then experimentally,with measurements in anechoic chamber. Atlast, in-situ measurements have been performed on awall corner made of concrete to assess the methodon a real case., Dans la situation d’un radar monostatique la configuration diédrique présente l’avantage de renvoyer un signal fort grâce au mécanisme de double rebond. De plus, il s’agit d’une configuration omniprésente par exemple en zone urbaine, avec des structures rue-bâtiment, ou en forêt, avec des structures de type sol-arbre. Les signaux rétrodiffusés sont donc liés à ces structures, plus précisément aux matériaux les composant, via leurs permittivités diélectriques. Retrouver ces permittivités depuis la mesure des champs rétrodiffusés permettrait l’identification des matériaux de la structure diédrique. Ceci mène à des applications telles que la détection de défauts dans la structure, l’estimation de la teneur en eau des sols ou du béton, ou in fine à des thématiques de classification, pour des objets dans des images radar par exemple, ou de détection de changement. En modélisant le mécanisme de double rebond lié à la structure de dièdre par deux réflexions successives via les coefficients de Fresnel, trois méthodes d’inversion des permittivités ont été mises en place au cours de la thèse. La première, analytique, utilise à un angle d’incidence donné la mesure des champs copolarisés. Les deux autres méthodes, alliant une partie analytique et une partie numérique, nécessitent la mesure, du ratio polarimétrique pour l’une, du champ horizontal pour l’autre, effectuée à deux angles d’incidence différents. Ces méthodes ont été validées numériquement à l’aide de la méthode d’Optique Géométrique fournie par le logiciel FEKO. Des mesures en chambre anéchoique ont été effectuées afin de valider ces méthodes. De même une mesure in-situ a été effectuée sur un dièdre en béton formé par un angle entre deux murs.

For a monostatic radar, the dihedral arrangementexhibits a strong response compare tobare surface. This phenomenon can be seen in radarimages of urban areas or forestry as building andstreets and trunks and ground are forming such structures.This signature can be modelled with the doublebounce equation as a cascade of Fresnel reflectioncoefficients. These coefficients depend on the permittivitiesof the structures. Finding these permittivitiesfrom the backscattered fields may allowed to identifythe materials involved in the dihedral. Thereforethiscan be applied for default detection, moisture retrievaland classification topics for instance.By modelling the double bounce scattering with Fresnelcoefficients, three methods have been developedto retrieve these permittivities. The first one uses asingle incidence angle and the measurements of copolarisedfields. The last two methods need the measurementof the copolarised fields ratio or of the horizontalelectric field alone at two different incidenceangles. All these methods have been validated, firstnumerically using FEKO RL-GO method, then experimentally,with measurements in anechoic chamber. Atlast, in-situ measurements have been performed on awall corner made of concrete to assess the methodon a real case.; Dans la situation d’un radar monostatique la configuration diédrique présente l’avantage de renvoyer un signal fort grâce au mécanisme de double rebond. De plus, il s’agit d’une configuration omniprésente par exemple en zone urbaine, avec des structures rue-bâtiment, ou en forêt, avec des structures de type sol-arbre. Les signaux rétrodiffusés sont donc liés à ces structures, plus précisément aux matériaux les composant, via leurs permittivités diélectriques. Retrouver ces permittivités depuis la mesure des champs rétrodiffusés permettrait l’identification des matériaux de la structure diédrique. Ceci mène à des applications telles que la détection de défauts dans la structure, l’estimation de la teneur en eau des sols ou du béton, ou in fine à des thématiques de classification, pour des objets dans des images radar par exemple, ou de détection de changement. En modélisant le mécanisme de double rebond lié à la structure de dièdre par deux réflexions successives via les coefficients de Fresnel, trois méthodes d’inversion des permittivités ont été mises en place au cours de la thèse. La première, analytique, utilise à un angle d’incidence donné la mesure des champs copolarisés. Les deux autres méthodes, alliant une partie analytique et une partie numérique, nécessitent la mesure, du ratio polarimétrique pour l’une, du champ horizontal pour l’autre, effectuée à deux angles d’incidence différents. Ces méthodes ont été validées numériquement à l’aide de la méthode d’Optique Géométrique fournie par le logiciel FEKO. Des mesures en chambre anéchoique ont été effectuées afin de valider ces méthodes. De même une mesure in-situ a été effectuée sur un dièdre en béton formé par un angle entre deux murs.

International audience; The joint estimation of means and scatter matrices is often a core problem in multivariate analysis. In order to overcome robustness issues, such as outliers from Gaussian assumption, M-estimators are now preferred to the traditional sample mean and sample covariance matrix. These estimators are well established and studied in the real case since the sev-enties. Their extension to the complex case has drawn recent interest. In this letter, we derive the asymptotic performance of complex M-estimators for multivariate location and scatter matrix estimation.

This paper deals with lower bounds on aperiodic correlation of sequences. It intends to solve two open questions. The first one is on the validity of the Levenshtein bound for a set of sequences other than binary sequences or those over the roots of unity. Although this result could be a priori extended to polyphase sequences, a formal demonstration is presented here, proving that it does actually hold for these sequences. The second open question is on the possibility to find a bound tighter than Welch’s, in the case of a set consisting of two sequences $M = 2$ . By including the specific structure of correlation sequences, a tighter lower bound is introduced for this case. Besides, this method also provides in the cases $M = 3$ and $M = 4$ a tighter bound than the up-to-now tightest bound provided by Liu et al.

International audience; This paper explores the problem of change detection in time series of heterogeneous multivariate synthetic aperture radar images. Classical change detection schemes have modeled the data as a realization of Gaussian random vectors and have derived statistical tests under this assumption. However, when considering high-resolution images, the heterogeneous behavior of the scatterers is not well described by a Gaussian model. In this paper, the data model is extended to spherically invariant random vectors where the heterogeneity of the images is accounted for through a deterministic texture parameter. Then, three separate detection prob- lems are considered and generalized likelihood ratio test technique is used to derive statistical tests for each problem. The constant false alarm rate property of the new statistics are studied both the- oretically and through simulation. Finally, the performance of the new statistics are studied both in simulation and on real synthetic aperture radar data and compared to Gaussian-derived ones. The study yields promising results when the data are heterogeneous.

International audience

International audience

In this work, a novel target detector for hyperspectral imagery is developed. The detector is independent on the unknown covariance matrix, behaves well in large dimensions, distributional free, invariant to atmospheric effects, and does not require a background dictionary to be constructed. Based on a modification of the robust principal component analysis (RPCA), a given hyperspectral image (HSI) is regarded as being made up of the sum of a low-rank background HSI and a sparse target HSI that contains the targets based on a pre-learned target dictionary specified by the user. The sparse component is directly used for the detection, that is, the targets are simply detected at the non-zero entries of the sparse target HSI. Hence, a novel target detector is developed, which is simply a sparse HSI generated automatically from the original HSI, but containing only the targets with the background is suppressed. The detector is evaluated on real experiments, and the results of which demonstrate its effectiveness for hyperspectral target detection especially when the targets are well matched to the surroundings., Comment: Accepted for publication in the book "Hyperspectral Image Analysis - Advances in Signal Processing and Machine Learning". arXiv admin note: text overlap with arXiv:1711.08970, arXiv:1808.06490

International audience

International audience; An experimental airborne passive radar system exploiting DVB-T signal has been developped and embedded on a motoglider dedicated to test flight. Two experimental campaigns have been run to acquire of real data. This article presents the level of the different contributors observed in the experimental data and analyses the bistatic clutter spread.

International audience; This paper proposes a new approximation of the theoretical Signal to Interference plus Noise Ratio (SINR) loss of the Low-Rank (LR) adaptive filter built on the eigenvalue decomposition of the sample covariance matrix. This new result is based on an analysis in the large dimensional regime, i.e. when the size and the number of data tend to infinity at the same rate. Compared to previous works, this new derivation allows to measure the quality of the adaptive filter near the LR contribution. Moreover, we propose a new LR adaptive filter and we also derive its SINR loss approximation in a large dimensional regime. We validate these results on a jamming application and test their robustness in a Multiple Input Multiple Output Space Time Adaptive Processing (MIMO-STAP) application where the data size is large

International audience; A preliminary Doppler analysis in the UHF-band on a man walking in free space is presented in this paper. Firstly, the back-scattered response of the moving target has been provided at 1 GHz by a simulation tool based on physical optics theory. Then, a short-time Fourier transform (STFT), a reassignment spectrogram (RE-Spect) and a Wigner-Ville distribution (WVD) have been employed to the data, in order to characterize the micro-Doppler signature of the human body walking. The results are based entirely on numerical tests. The investigation on the time variations of the Doppler spectrum of moving targets in UHF-band is of interest for the emerging radar applications devoted to the detection and tracking of people in highly cluttered environment.

International audience; This paper is interested in phase code design, and especially in their robustness to the eclipsing phenomenon. In a radar background, the term "eclipsing" describes a partial loss of data at the reception side. It implies an increase of pulse compression sidelobes, and thus a degradation of detection performance. In this regard, this paper describes an optimisation procedure which aims to provide waveforms (phase codes) that are quite robust to the eclipsing phenomenon. Some simulations are provided to highlight the usefulness of the procedure.

International audience; Deploying neural networks models over embedded devices have an increased interest and many works is ongoing on that topic. Energy consumption, model sizes and inference time are critical issues as explained in the literature. In the context of IoT and edge computing, tradeoff have been studied in order to get a low cost but rapid answer, robust to connection issue exploiting early exiting or distributing deep neural networks. Those approaches exploits the cloud as an endpoint, balancing the load with respect to different computing capabilities. In this paper, we propose to extend those approaches to networks of embedded devices such as a swarm of drones, where every device has the same computing capabilities (in terms of energy and speed). Computing load may be balanced among the whole swarm in order to maximise either the lifetime of specific devices or lifetime of the whole swarm. We develop criteria to best cut and distribute those networks, validate them through power measurement and express the different tradeoffs we have to address.

International audience

International audience; Detecting targets embedded in a noisy environment is an important topic in adaptive array processing. In the traditional statistical framework, this problem is addressed through a binary hypothesis test, which usually requires the estimation of side parameters from secondary data. The latter are assumed to be homogeneous and target-free, which is in practice questionable. Indeed, secondary data are usually corrupted by radar clutters and/or jammers which can be non-stationary and locally low rank. Fortunately, the latter behaviors can be well acknowledged by a union-of-subspaces model. In this work, we propose a modified subspace clustering model which can be solved using convex optimization algorithms. In the context of multiple sparse target detection and localization, a comparison is performed with various robust detection methods exhibiting advantages and drawbacks of the proposed one.

In most modern signal processing applications, observations are generally modeled by non-Gaussian distributions with covariance matrices exhibiting a particular structure. Taking these properties into account in the estimation scheme improves drastically the estimation accuracy. In this paper, we consider the estimation of structured scatter matrix, where the assumed model can differ from the true model of the data. Specifically, we propose a novel class of estimators, named StructurEd ScAtter Matrix Estimator (SESAME) in the mismatched framework. We also conduct a theoretical analysis of its asymptotic performance., Dans la plupart des méthodes récentes en traitement du signal, les données sont modélisées par des distributions non gaussiennes, dont la matrice de covariance possède une structure particulière. Prendre ces propriétés en compte dans le processus d'estimation améliore nettement la qualité des estimées. Dans ce papier, nous considérons l'estimation de matrices de dispersion structurées, où le modèle supposé peut différer du vrai modèle des données. Plus précisément, nous proposons une nouvelle classe d'estimateurs dénommée StructurEd ScAtter Matrix Estimator (SESAME) dans un contexte de modèles mal spécifiés. Nous menons aussi une analyse théorique de leurs performances asymptotiques.

Passive radar detection benefits from non-cooperative telecommunication broadcasters, already existing in the environment, such as DVB-T broadcasters.It uses signal reflections over potential targets. An airborne receiver takes advantage of it as radar echoes.This new kind a detection system is discrete, has low energy consumption, uses already allocated frequencies and broaden radar detection to low altitudes.Due to airborne constraints, the standard signal processing steps, as the receiving system need to be adjusted.Indeed a dedicated radar has been developped in order to get experimental signal, and therefore help deepen the understanding of propagation phenomenon.The proposed reference signal estimation allows to face channel multipath induced fluctuations on the one hand, and to take into account channel time variationsthanks to Basis Expansion Model (BEM) modeling. A theoretical analysis of the clutter spread is then drawn.Experimental results confirm the expectation in the range-Doppler and angle-Doppler domain.Besides a clutter cartesian projection highlights the major reflectors, that may be confronted to the terrain truth.Reference signal estimation and clutter spread constitute two radar detection pilars, as these components have to be cancelled.So as to reject direct path, space filtering orthogonal to the direct direction is also performed to suppress the impact of the transmitted noise.Then reference signal is cancelled via a standard rejection method based on least-square filter coefficients estimation.The large Doppler and range clutter spread, lead us to reject the reference signal over shorter correlation periods.The present work gives an accurate comprehension of propagation mechanisms impact on airborne passive radar signal processing andprovides a promising perspective regarding intermediate radar cross section target detection.; La détection radar passive met à profit des émetteurs non-coopératifs, déjà présents dans l’environnement, qui transmettent des signaux de télécommunications, de type DVB-T dans l’étude présentée.Elle utilise les réflexions de ces signaux sur de potentielles cibles et les exploite comme échos radar au niveau d’un récepteur aéroporté.Ces nouveaux systèmes de détection, par nature discrets et économes en énergie et en allocation de fréquences, étendent la surveillance à la basse altitude.Si les différentes étapes des traitements classiques utilisés en radar passif terrestre (estimation du signal de référence, réjection, filtrage adapté, détection)demandent d’être réorientées sérieusement pour répondre aux contraintes liées à la réception aéroportée,il en va de même du récepteur qui doit satisfaire les exigences matérielles de la plateforme aérienne.Dans ce but, un système expérimental embarqué sur motoplanneur a été développé permettant d’acquérir des signaux réels indispensablesà la compréhension de l’impact de la propagation des signaux DVB-T.La méthode d’estimation du signal de référence proposée permet d’une part, de lutter contre les fluctuations du canal de propagation induites par les multi-trajetsen exploitant la diversité d’antenne et d’autre part, de prendre en compte les variations temporelles en s’appuyant sur la méthode BEM (Basis Expension Model).Ensuite, une analyse théorique sur la répartition du fouillis de sol est apportée.L’exploitation des signaux expérimentaux permet de la valider par une analyse dans le plan distance-Doppler et angle-Doppler.Une projection cartésienne permet de mettre en évidence des échos forts confrontés avec la vérité terrain.L’estimation du signal de référence et la connaissance de l’étalement du fouillis de sol sont les piliers fondamentaux de la détection car ces composantes représentent deux contributions à rejeter.Pour le signal de référence, une méthode classique de réjection où les coefficients du filtre sont estimés au sens des moindres carrés est mise en oeuvre.Un filtrage spatial orthogonal à la direction d’arrivée du signal de référence est ajouté afin de diminuer l’impact du bruit émis.Le large étalement en Doppler et en distance nous a conduit à rejeter le fouillis sur des périodes de corrélation plus courtes.Les travaux présentés apportent une compréhension fine de l’impact de la propagation sur les traitements de détection en radar passif aéroporté et offrent des perspectives engageantesquant à la détection de cibles de moyennes à grandes Surfaces Equivalentes Radar.

International audience; We report that metamaterial-inspired one-dimensional gratings (or metagratings) can be used to control nonpropagating diffraction orders as well as propagating ones. By accurate engineering of the near field, it becomes possible to satisfy power-conservation conditions and achieve perfect control over all propagating diffraction orders with passive and lossless metagratings. We show that each propagating diffraction order requires 2 degrees of freedom represented by passive and lossless loaded thin "wires." This provides a solution to the old problem of power management between diffraction orders created by a grating. The theory developed is verified by both three-dimensional full-wave numerical simulations and experimental measurements, and can be readily applied to the design of wavefront-manipulation devices over the entire electromagnetic spectrum as well as in different fields of physics.

International audience; This paper addresses the problem of the clutter subspace projector estimation in the context of a disturbance composed of a low rank heterogeneous (Compound Gaussian) clutter and white Gaussian noise. In such a context, adaptive processing based on an estimated orthogonal projector onto the clutter subspace (instead of an estimated covariance matrix) requires less samples than classical methods. The clutter subspace estimate is usually derived from the eigenvalue decomposition of a covariance matrix estimate. However, it has been previously shown that a direct maximum likelihood estimator of the clutter subspace projector can be obtained for the considered context. In this paper, we derive two algorithms based on the block majorization-minimization framework to reach this estimator. These algorithms are shown to be computationally faster than the state of the art, with guaranteed convergence. Finally, the performance of the related estimators is illustrated on realistic Space Time Adaptive Processing for airborne radar simulations.

International audience; In an earlier work the design of a 2:1 bandwidth dual polarized phased array of Archimedean spiral antennas with cavity was presented. The reflection coefficient, circular polarization and grating lobes goals were obtained independently in the design process. The low end of the final bandwidth was constrained to the frequency cutoff of the spiral antenna, at the low end of its bandwidth. In the present paper, an extension of the bandwidth of the array is achieved by connecting the spiral antennas of the array. Hence, with the same size, the bandwidth is enlarged almost 3 times (a total of 6:1). The measurements confirm the simulations. The array can scan up to = 30 ??? keeping a low reflection coefficient, good right and left hand circular polarization, low side lobe levels and using a low profile cavity.

International audience; The spherically invariant random process (SIRP) clutter model is commonly used in scenarios where the radar clutter cannot be correctly modeled as a Gaussian process. In this short communication, we devise a novel Maximum-Likelihood (ML)-based iterative estimator for direction-of-departure and direction-of-arrival estimation in the Multiple-input multiple-output (MIMO) radar context in the presence of SIRP clutter. The proposed estimator employs a stepwise numerical concentration approach w.r.t. the objective function related to the marginal likelihood of the observation data. Our estimator leads to superior performance, as our simulations show, w.r.t. to the existing likelihood based methods, namely, the conventional, the conditional and the joint likelihood based estimators, and w.r.t. the robust subspace decomposition based methods. Finally, interconnections and comparison between the Iterative Marginal ML Estimator (IMMLE), Iterative Joint ML Estimator (IJMLE) and Iterative Conditional ML Estimator (ICdMLE) are provided.

International audience; A planar array of symmetrical square spiral antennas over a ground plane have resonances when scanning off-broadside. This paper demonstrates that resonances also appear in a single symmetrical receiving spiral antenna without a ground plane, provided that the incident wave is off-broadside. Transmission line analysis predicts the resonances.