DLR Portal
Home
|
Sitemap
|
Kontakt
|
Barrierefreiheit
Impressum und Nutzungsbedingungen
Datenschutz
Cookies & Tracking
|
English
Sie sind hier:
Home
:
Aktuelles
:
Wissenschaftliche Kolloquien
:
Institutsseminar 2019
Erweiterte Suche
Aktuelles
Wissenschaftliche Kolloquien
Standortkolloquium
Institutsseminar
Institutsseminar 2024
Institutsseminar 2023
Institutsseminar 2022
Institutsseminar 2021
Institutsseminar 2020
Institutsseminar 2019
Institutsseminar 2018
Institutsseminar 2017
Institutsseminar 2016
Institutsseminar 2015
Institutsseminar 2014
Institutsseminar 2013
Institutsseminar 2012
Institutsseminar 2011
Institutsseminar 2010
Institutsseminar 2009
Institutsseminar 2008
Institutsseminar 2007
Institut
Abteilungen
Compact Test Range
Projekte
TanDEM-X Science
Tandem-L Science
Helmholtz-Allianz
Publikationen des Instituts
Preisverleihungen
Stellenangebote
Angebote zu Praktikum, Bachelor-, Masterarbeiten und Dissertationen
Angebote für Schüler und Studenten
Archiv - Konferenzen
Software
Anreise
Externe Besucher werden gebeten unsere Seminarorganisation zu kontaktieren (
siehe hier
)
.
Vorträge im Institutsseminar 2019
Nestor Yague-Martinez "Burst-Mode Wide-Swath SAR Interferometry for Solid Earth Monitoring"
Montag, 9. Dezember 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
This work addresses the generation of large area maps of ground deformation originated by dynamic processes of the Earth using spaceborne Synthetic Aperture Radar (SAR) systems following a burst-mode acquisition strategy. The main focus lies on the development of new interferometric processing techniques for burst-mode SAR data and on the proposal, design, experimental implementation and demonstration of an acquisition mode with wide-swath capabilities, able to improve the accuracy of the measurements. The Terrain Observation by Progressive Scans (TOPS) mode achieves wide-swath coverage by recording bursts of echoes over different subswaths. Additionally, a steering of the antenna beam in azimuth is performed in order to equalize the signal-to-noise ratio (SNR) and azimuth ambiguitiy-to-signal ratio (AASR), i.e., avoiding the so-called “scalloping” effect, inherent to the ScanSAR mode. The first contribution of this work consists on the consolidation of the interferometric processing algorithms of pairs of Sentinel-1 TOPS images with a focus on the methodology to fulfill the stringent azimuth coregistration requirements. The second contribution provides a methodology for coregistering stacks of Sentinel-1 TOPS acquisitions exploiting a joint approach. Due to the quasi-polar orbit configuration of remote sensing satellites, the sensitivity to the North-South (N-S) direction remains low, since radar systems provide measurements in the line-of-sight direction, perpendicular to along-track. In the third contribution, correlation techniques with TerraSAR-X StripMap data are applied to the Mw9.0 Tohoku-Oki Earthquake occurred in 2011 in Japan, which allows obtaining 2-D deformation maps, i.e., in line-of-sight and along-track directions. In case of using TOPS data, correlation techniques provide a poor performance in the along-track direction due to its low azimuth resolution. The fourth contribution intends to overcome this limitation by proposing the novel 2-look TOPS acquisition mode, a wide-swath mode, able to achieve sensitivity to azimuth displacements by exploiting spectral diversity techniques. A demonstration with experimental TerraSAR-X data for the mapping of fast displacements with pairs of images and for the mapping of slow displacements with time series shows its high potential for solid Earth monitoring.
Georg Fischer "Modeling of Subsurface Scattering from Ice Sheets for Pol-InSAR Applications"
Freitag, 22. November 2019
11.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
Polarimetric synthetic aperture radar interferometry (Pol-InSAR) measurements are known since decades to depend on the geophysical properties below the surface of ice sheets, because of the penetration of microwave signals into dry snow, firn, and ice, but only very few studies have addressed this topic. Two potential Pol-InSAR applications are the compensation of the penetration bias in InSAR DEMs and the estimation of geophysical properties of the subsurface of glaciers and ice sheets. However, the required models to link the Pol-InSAR measurements to the subsurface properties are not yet established. The aim of this thesis is to improve the modeling of the vertical backscattering distribution in the subsurface of ice sheets and its effect on polarimetric interferometric SAR measurements at different frequencies. In order to achieve this, airborne SAR data from two different test sites on the Greenland ice sheet are investigated. This thesis contributes with three concepts to a better understanding and to a more accurate modeling of the vertical backscattering distribution in the subsurface of ice sheets. First, the integration of scattering from distinct subsurface layers. These are formed by refrozen melt water and cause an interesting coherence undulation pattern, which cannot be explained with previously existing models. The second step is the improved modeling of the general vertical backscattering distribution of the subsurface volume. The advantages of more flexible volume models are demonstrated, but interestingly, the simple modification of a previously existing model with a vertical shift parameter lead to the best agreement between model and data. The third contribution is the model based compensation of the penetration bias, which is experimentally validated. This thesis therefore improves the state of the art of subsurface scattering modeling for Pol-InSAR applications, demonstrates the model-based penetration bias compensation, and makes a further research step towards the retrieval of geophysical subsurface information with Pol-InSAR.
Paul A. Rosen "An Update on the NASA-ISRO SAR Mission and an Architecture for its Successor"
Dienstag, 12. November 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
With the expected launch of the NASA-ISRO SAR (NISAR) Mission in 2022, NASA will have a free-flying SAR in space for the first time since the 1978 SeaSAT mission. NISAR has been designed to acquire long-anticipated measurements for a broad scientific constituency – geophysicists, glaciologists, ecologists and scientists in a range of related disciplines – with an expectation that the data will be useful for operational users and applications developers. As such, the observation and product dissemination plans strive to balance science and applications performance across disciplines by comprehensively and systematically covering Earth’s land and ice-covered surfaces twice (on ascending and descending passes) each 12-day repeat cycle in only a handful of consistently repeated radar modes. This talk will describe the NISAR observatory, its design, science objectives, current status, expected products, and will address some of the important trades and operational choices needed to achieve scientific balance. These experiences also have an impact on NASA’s plans for SAR continuity beyond NISAR, currently being studied as part of NASA’s Decadal Survey Designated Observables program.
Victor Cazcarra Bes "Forest Structure Characterization from SAR Tomography"
Montag, 11. November 2019
11.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
The study of forest structure is required for many forest applications like biomass estimation, forest productivity, plant biodiversity or forest management. Synthetic Aperture Radar Tomography (TomoSAR) allows a direct 3D imaging of the illuminated scene. Furthermore, the future tomographic SAR missions (like Biomass or Tandem-L) will use this technique to provide a 3D representation of the forests at a global scale. In this context, the Ph.D. has the main goal of linking forest structure with TomoSAR data. In the first part of the thesis, a framework that allows the qualitative and quantitative characterization of the 3-D forest structure from tomographic SAR profiles has been proposed. From the statistical distribution of the local maxima of the TomoSAR profiles, two indices are proposed to describe the horizontal density and the vertical complexity of a forest. This approach has been evaluated on a TomoSAR L-band data set over a temperate forest in Germany, together with the structure obtained from single-tree ground measurements and Lidar data. In the second part of the thesis, different TomoSAR algorithms and configurations with a focus on forest structure applications have been evaluated. Fourier beamforming, Capon beamforming and compressive sensing have been analyzed with simulated and real TomoSAR data at L-band. Finally, the requirements of TomoSAR acquisitions for distinguishing different structure types have been evaluated in terms of vertical resolution, height of ambiguity and peak side-lobe level of the point spread function. The results of the thesis underline the potential of TomoSAR as a 3-D imaging tool for forest structure applications and contribute to better understand the TomoSAR profiles. Moreover, they open the door to further evaluate, investigate and exploit the enormous amount of TomoSAR data that the two upcoming SAR missions (ESA’s BIOMASS and DLR’s Tandem-L) will provide over all forests of the Earth.
Michele Martone "Onboard Quantization for Interferometric and Multichannel SAR Systems"
Mittwoch, 6. November 2019
14.00 h Besprechungsraum HR - 024, Gebäude 102
Abstract:
Synthetic aperture radar (SAR) systems allow for all-weather, day-night, high-resolution imaging and are therefore very attractive for a broad range of scientific and commercial applications related to Earth and planetary observation. For present and next-generation spaceborne SAR missions, an increasing volume of onboard data is going to be required, due to the employment of large bandwidths, multiple channels, and the imaging of large swath widths at fine spatial resolutions. This implies strong requirements in terms of onboard memory and downlink capacity, which are limited resources and often represent a bottleneck in the design of SAR missions. In this context, the proper digitization of the raw data acquired by the SAR sensor represents an aspect of utmost importance, since the number of bits allocated for the recorded radar signal defines the total volume of data to be managed by the system and, at the same time, it directly affects the performance of the resulting SAR products. In this thesis, the impact of quantization on SAR and interferometric (InSAR) performance has been investigated using experimental data acquired by the X-band SAR mission TanDEM-X for different system parameters and SAR scene characteristics. By this, a performance-optimized quantization is introduced, which allows for a joint optimization of the data rate and of the interferometric performance, by exploiting the a priori knowledge about the SAR backscatter information. Furthermore, a novel azimuth-switched quantization (ASQ) for the efficient implementation of non-integer quantization rates has been developed. The proposed technique has been demonstrated on TanDEM-X experimental data and grants high flexibility in terms of performance and resource allocation. The second part of the dissertation focuses on the definition of innovative onboard data reduction methods for next-generation SAR systems, such as multi-azimuth channel (MAC) SAR and staggered SAR. For such systems, a significant oversampling and consequent correlation properties of the azimuth SAR raw signal is exploited by applying an efficient encoding and quantization of the SAR raw data. In this way, a significant reduction of the data volume is achieved at the cost of a modest increase of onboard computational effort.
Philipp Posovszky und Sebastian Iff "Introduction and migration to DLR GitLab"
Dienstag, 5. November 2019
14.00 h Besprechungsraum HR-024, Gebäude 102
Abstract:
Since the end of September the central IT-Service „DLR-GitLab“ is available. It is the successor of the version control system Subversion (SVN) Furthermore it is a platform for collaborative software development that is accessible for all DLR employees as well as for external partners. As already announced it is planned that the DLR-GitLab service supersedes the HR-GitLab. In this HR-Seminar we would like to give a brief overview of the DLR-GitLab as well as “good practice” tips. We also invite everyone to express and discuss concerns about the migration from the HR-GitLab to the DLR-GitLab.
Andrea Pulella and Rodrigo Aragao Santos "Forest mapping from Sentinel-1 InSAR data"
Dienstag, 29. Oktober 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
The study of global forest coverage and biomass is fundamental for assessing forests’ impact on the ecosystem. After the recent deforestation acts in the Amazon’s state, forest monitoring is a real hot topic and remote sensing represents a powerful tool for an effective response to this problem. The Copernicus space program, thanks to its ‘Sentinel’ satellites provides a large amount of data from optical to Synthetic Aperture Radar (SAR) sensors. The Sentinel-1 (S1) satellites operate SAR sensors and provide data with short revisit-time that are interferometric compatible. The proposed talk will introduce a novel methodology for the exploitation of S1 Interferometric wide-swath (IW) acquisitions for land cover classification. Indeed, by observing the temporal trend of the interferometric coherence, reliable information about the nature of the observed target can be retrieved and eventually used as input to machine learning-based classifiers. The case study of the Amazon rainforest is considered and, in the specific, a framework for the mapping of forested areas over the Rondonia state is presented. Eventually, a strategy for the systematic monitoring of world’s forests will be discussed together with the open challenges and opportunities.
Maxwell Nogueira Peixoto "Nadir Echo in Staggered SAR: Characterization and Processing for Suppression"
Montag, 28. Oktober 2019
10.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
High-resolution wide-swath SAR systems can be achieved by the simultaneous imaging of multiple sub-swaths through digital beamforming. In conventional SAR systems, these sub-swaths are separated by blind ranges from which no data is acquired. Staggered SAR is a novel SAR operation mode characterized by a continuous variation of the pulse repetition interval, causing the blind ranges to move from pulse to pulse. With sufficient oversampling in azimuth, the blind ranges can be interpolated over and a wide continuous swath can be imaged with high resolution. Staggered SAR is the baseline acquisition mode of Tandem-L and under consideration for NISAR and the future missions of ESA’s Copernicus Programme. In conventional SAR systems, nadir echoes appear at bright strips at specific ranges and are avoided by the use of timing diagrams to align these echoes with the blind ranges. This approach is not possible in staggered SAR, as both nadir echoes and blind ranges move from pulse to pulse. Consequently, nadir echoes are present in staggered SAR images, which raises the questions of what is the impact of these echoes and what can be done to suppress them. In this work, a parametric model of nadir echo based on a dedicated TerraSAR-X acquisition was developed. This model, along with a theoretical description of the positioning of nadir echoes in staggered SAR data, is used to assess the effect of nadir echoes in staggered SAR images through simulations based on real TerraSAR-X data. Additionally, a novel processing technique capable of suppressing the nadir echoes is proposed and its performance is evaluated. The results of this work extend the staggered SAR theory and can be used to evaluate the impact of nadir echo in a specific staggered SAR system or to impose requirements on the nadir attenuation provided by the antenna pattern.
André Barros Cardoso da Silva "A Priori Knowledge-Based Post-Doppler STAP for Traffic Monitoring with Airborne Radar"
Freitag, 18. Oktober 2019
9.30 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
Traffic monitoring is currently a trending topic due to the worldwide increase of road users. Safe and efficient roadway operations require detailed traffic information that relies on stationary ground infrastructure often available only for the major highways. Particularly in case of catastrophes (when mobile internet and phone communication are impossible), the actual road systems could fail altogether due to extensive power blackouts or ground infrastructure damages, resulting in complete lack of information. Synthetic aperture radar (SAR) offers a remarkable solution for this scenario due to its unique capability in providing high-resolution images independent of daylight and weather conditions, allowing applications in ground moving target indication (GMTI). State-of-the-art GMTI algorithms have been proposed in the literature using a priori knowledge information, whereas their robustness is often achieved with high costs, high hardware complexity and high computational effort. This doctoral thesis presents a GMTI processor based on the powerful multi-channel technique post-Doppler space-time adaptive processing (PD STAP). This GMTI processor blends a road map with a digital elevation model in order to recognize and to reject false detections, and to assign the detected vehicles to their correct positions on the roads. Accurate position, velocity and moving direction estimates of the vehicles are obtained with decreased processing hardware complexity and low costs compared to state-of-the-art systems. This doctoral thesis also includes an efficient data calibration algorithm that corrects the channel imbalances and the Doppler centroid variations over range and azimuth. In addition, novel and automatic training data selection strategies are presented for the clutter covariance matrix estimation, which is essential for the PD STAP processor, since it impacts directly its clutter cancellation and target detection capabilities. The novel PD STAP processor has three operational modes designed for military and civilian applications, including a fast processing mode that paves the way for real-time traffic monitoring. All modes were tested on multi-channel data acquired by DLR’s airborne system F-SAR, containing scenarios with controlled vehicles and real-traffic.
Tobias Bollian "Digital Beamforming for Radio Frequency Interference Suppression in Synthetic Aperture Radar"
Donnerstag, 17. Oktober 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
Synthetic Aperture Radar (SAR) has developed into a powerful tool for the frequent Earth observation of large areas, independent of the weather and sunlight conditions. However, the contamination of the acquired SAR data by Radio Frequency Interference (RFI) has become an increasing problem. This observed trend is expected to worsen in the future due to the growing demand for other wireless services. As the SAR center frequency is determined by the physical properties that are to be retrieved, switching frequency bands is not an option. Effective methods to mitigate RFI are therefore critical for the successful operation of a SAR mission. In a conventional SAR, RFI correction relies on notching methods that also remove parts of the SAR signal. Otherwise, a-priori knowledge about the interfering signal characteristics is required and the filtering performance strongly depends on the observed scene and interference. New, advanced SAR instruments employ a multi-channel architecture that is capable of Digital Beamforming (DBF). These systems can spatially filter the received signals and give rise to new RFI mitigation techniques. The antenna pattern can be adaptively set after data acquisition to null the angular direction of the impinging RFI signal. Even more, DBF gives the opportunity to individually measure RFI and SAR signals arriving from different directions. This allows the gathering of information about the interfering signal characteristics from the data. This work proposes novel DBF-based RFI mitigation techniques that utilize the spatial distribution of the SAR signal in the range-Doppler domain. The spatial distribution is variable in time and yet predictable because the arrival time of the radar return is determined by the imaging geometry. The proposed new methods can be divided into two categories. The first new approach adaptively nulls the antenna pattern throughout the SAR data to suppress the RFI signal. A simulator was implemented in the framework of this thesis to verify the performance and a proof-of-concept with experimental airborne data is presented. The second new approach steers simultaneous digital beams into the direction of the interferers to measure and subtract them from the data. Further, the measured interferer signal characteristics can then be used to estimate the RFI signal for time instances when a spatial filtering is not possible. Simulation results for this approach are presented.
Lucas Barroso Knupp"A hierarchical approach to airborne SAR motion compensation and antenna pattern correction"
Dienstag, 15. Oktober 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
Airborne SAR data focusing involves several processing steps that introduce signal corrections which vary both as a function of time and as a function of frequency. Prominent examples are the correction of non-linear sensor platform motion (motion compensation) and the correction of the propagation direction and frequency variant SAR antenna illumination diagram. Introducing time- and frequency-variant corrections necessarily involves a trade-off in time-domain versus spectral resolution. The seminar introduces a new and general approach to improve upon state-of-the-art algorithms by making a succession of incremental corrections at a series of different time/frequency resolutions. The approach is hierarchical and features a tree-like subdivision of the signal spectrum. The approach is designed to improve upon current state-of-the-art algorithms that use short-time-Fourier-Transforms. The aim is to better accommodate high frequency motion errors, sensor attitude angle changes or rapidly changing topographic terrain. The presentation explains and illustrates existing techniques and uses examples to motivate the new development. Algorithms are evaluated and compared on the basis of simulated raw data corresponding to real airborne SAR acquisition geometries.
Andrés Felipe Betancourt Payan "Towards the Performance Assessment of a Mapdrift Autofocus for a P-Band SAR Mission Implementation"
Montag, 23. September 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract: In the context of the ESA BIOMASS mission, for the first time, a P-Band SAR sensor is going to be mounted into a spaceborne system. With its penetration capability, it will contribute to the measurement of the biomass and carbon content in the Earth's forests. An autofocus algorithm is needed for the correction of the phase errors introduced by the changing refractive index in the ionosphere. Because of the quickly changing nature of the ionosphere, defocusing has to be measure and corrected locally over several sections of a SAR capture.
Jan Reifenhäuser "Qualifizierung einer gedruckten Ka-Band Hohlleiterschlitzantenne im Temperaturbereich von 50 bis -40°C"
Dienstag, 17. September 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
Zunehmend kleinere Strukturen bei der Entwicklung von Antennen in hohen Frequenzbereichen stellen hohe Ansprüche an die Fertigungsmethoden. Eine Alternative zu den bisher gängigen subtraktive Fertigungsverfahren stellt das 3D Druckverfahren in Kunststoff dar. Die Materialeigenschaften des Kunststoffes für die Anwendungen in Antennen sind jedoch weitestgehend unbekannt. Das Seminar behandelt die Untersuchung einer 3D gedruckten Ka-Band Hohlleiterschlitzantenne in den für Flugzeug SAR Anwendungen zu erwartenden Temperaturen. Der Fokus liegt dabei vor allem auf der messtechnischen Untersuchung einer Antenne innerhalb eines Klimaschranks. Dabei wird eine mit verhältnismäßig einfachen Mitteln zu realisierende Messanordnung und simple Methoden zur Verbesserung der Daten vorgestellt.
Dr. Nicholas Rattenbury, Dr. John Cater, Dr. Andrew Austin, University of Auckland, New Zealand "Ātea Zusammenarbeit: New Zealand/Germany Space research"
Dienstag, 10. September 2019
14.00 h Besprechungsraum 024 HR, Gebäude 102
Abstract:
New Zealand recently joined the exclusive club of nations that has sovereign access to space. Our current and recent work at The University of Auckland has been to ensure that we are preparing our students to contribute to the space economy both locally and abroad. We will describe our current and planned space science and engineering teaching and research programmes, with particular emphasis on the opportunities that we see arising from collaborating with German institutions. We have research programmes in synthetic aperture radar, deployable structures, low delta-v satellite propulsion and orbital mechanics and novel material science for heat shielding. We have interests in space-based earth observation, remote sensing and satellite optical communications. Our work now sits within the recently created Te Pūnaha Ātea -- Auckland Space Institute.
Josué Antonio López Ruiz "Tensor Decomposition and Deep Learning Neural Networks for Multispectral Remote Sensing Image Compression and Semantic Segmentation"
Montag, 22. Juli 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
Remote Sensing (RS) Multispectral Images (MSI) are nowadays very useful for classification of different materials in a scene of interest over the earth surface. The most recent used methods in the state of the art (SoA) for classification are those based on Machine Learning (ML). In recent years Deep Learning (DL) Artificial Neural Networks (ANN) have emerged with high accuracy classification algorithms. For image classification, as well as for semantic segmentation, Convolutional Neural Networks (CNN) have proven to be one of the highest accuracy methods for these tasks. Nevertheless, processing multispectral data in ANNs produces an extremely high number of computations and large processing time. In this work we propose a dimensionality reduction preprocessing stage for multispectral data, preserving statistical information and reducing data redundancy using Tensor Decomposition (TD) methods. This will attain high accuracy in CNNs for semantic segmentation in RS-MSIs but reducing in a large way computational complexity.
Marius Engel "Entwicklung und Aufbau eines UHF-Breitbandradars als drohnenbasiertes Sensorsystem"
Dienstag, 2. Juli 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
Radarsysteme sind in der Lage vergrabene Objekte im Erdreich zu detektieren. Ein Anwendungsgebiet stellt dabei die Detektion von Landminen und Blindgängern dar, in der Bedrohungen nahe der Erdoberfläche detektiert werden sollen. Im Rahmen des von 2012 bis 2016 stattgefundenen EU-Projekts TIRAMISU (Toolbox Implementation for Removal of Anti-personnel Mines, Submunitions and UXO), wurde am Institut für Hochfrequenztechnik und Radarsysteme ein neuartiges Mehrkanalsystem entwickelt, das kontaminierte Bereiche großflächig und zugleich aus sicherer Distanz untersuchen kann. Der nächste Schritt ist konsequenterweise, das Prinzip des TIRAMI-SAR auf einer drohnengestützten Plattform zu realisieren. Die Vorteile eines solchen Systems sind vor allem die luftgetragene Arbeitsweise des Radarsensors und die autonome Investigation vorgegebener Gebiete mit Hilfe GPS-gesteuerter Drohnen.
Das Seminar behandelt die theoretische Systemauslegung des entwickelten FMCW-Radars sowie die Evaluierung der einzelnen Teilkomponenten. Weiterhin werden die ersten Messergebnisse von SAR-Messungen mit Hilfe einer Linearachse diskutiert, die unter Laborbedingungen aufgenommen wurden.
Prof. Héctor Esteban González und Prof. Vicente Boria Esbert “Microwave Applications Group of the Polytechnic University of Valencia: presentation, research lines and activity in the field of Substrate Integrated Waveguides (SIW)”
Montag, 1. Juli 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
The Microwave Applications Group (GAM) of the Polytechnic University of Valencia (Spain) focuses its research activity on the analysis and design of passive components for high frequency applications implemented in various technologies (coaxial, waveguide, planar and substrate integrated waveguides), using conductive materials, dielectrics, periodic structures (EBGs) and tuning elements. These passive components are designed for being used in a wide range of communication systems (space, mobile and wireless communications, RADAR systems, and defense, remote sensing and radio navigation applications). The activity of the group began in 1997, and is currently made up of 15 people, including researchers and support staff. Additionally, there is a fluid collaboration with other universities both national and international, as well as with important companies in the Space sector, as well as the close relationship with the European Space Agency (ESA) and the Valencian Space Consortium (VSC). Substrate integrated waveguides (SIW) appeared in 2001 as a compromise between classical planar technologies and rectangular waveguides. It consists on a rectangular waveguide synthesized in a planar circuit board, so that it can be connected to other circuits in the same board. It can be manufactured with the same low cost manufacturing techniques as planar circuits, it is low profile and low weight, but it has lower losses and higher quality factor than planar circuits. However, it is still far from the high quality performance of the bulky and heavy rectangular waveguides, mainly due to the propagation of the waves through the lossy dielectric. In 2014 in the GAM proposed a new concept of Empty Substrate Integrated Waveguide (ESIW), which is a SIW where the dielectric has been removed. Since then, many high quality passive devices have been successfully implemented in ESIW (antennas, filters, couplers, phase shifters, circulators, …), obtaining performances close to that of the classical rectangular waveguide. The ESIW technology has also been tested for space conditions, proving to be a good alternative to rectangular waveguides for small satellites where size and weight are critical.
Gustavo Martin del Campo Becerra "On Descriptive and Statistical Regularization with TomoSAR Imaging as Application"
Dienstag, 18. Juni 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
The seminar treats the mathematical background of the most prominent regularization approaches adapted to solve the ill-conditioned nonlinear TomoSAR inverse problem, as an alternative to the commonly performed focusing methods, such as Capon beamforming and compressed sensing (CS). These techniques are assembled into two main groups: (i) deterministic descriptive regularization techniques, inspired by the Tikhonov’s regularization theory; and (ii) statistical regularization methods, in the context of the statistical decision-making theory. The differences and similarities between each regularization approach are to be discussed, as well as their main advantages and disadvantages. Taking into consideration the preceding addressed developments, several non-parametric approaches, based on maximum-likelihood (ML) and the weighted covariance fitting (WCF) criterion are introduced.
Jens Fischer „The Poisson Summation Formula and its role in radar“
Dienstag, 28. Mai 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
Very well-known in electrical engineering is the fact that sampling in one domain (time or frequency) results in periodization in the other domain (frequency or time). This simple circumstance is reflected in Poisson’s summation formula. However, Poisson’s summation formula is known in many, very different variants and it even depends on the actually used Fourier transform definition. This talk will give an overview over different variants of Poisson’s summation formula and how to interpret them. It is shown that by introducing two novel symbols, one for discretization (sampling) and one for periodization, any variant of a Poisson Summation Formula (PSF) reduces to a simple, well-known statement. The introduction of these two symbols moreover leads to a simple symbolic calculation scheme with a great potential for radar applications.
Victor Cazcarra Bes „Tomographic SAR algorithms for forest structure applications“
Donnerstag, 23. Mai 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
The study of forest structure is required for many forest applications like biomass estimation, forest productivity, plant biodiversity or forest management. The classical use of 2D measures to represent 3D structures, like a forest, has an inevitable loss of information. In front of this need, Synthetic Aperture Radar Tomography (TomoSAR) allows a direct 3D imaging of the illuminated scene. Furthermore, the future tomographic SAR missions (like Biomass or Tandem-L) will use this technique to provide a 3D representation of the forests at a global scale. In this context, the Ph.D. has the main goal of linking forest structure with TomoSAR data. In the first part of the seminar, descriptors based on TomoSAR data to quantitatively describe the 3D forest structure will be presented. These descriptors are validated in accordance with ecological definitions of forest structure based on individual tree measures as well as Lidar data. Moreover, as an application example, the monitoring of forest structure changes from TomoSAR over the Traunstein forest (Germany) using L-band data acquired by the F-SAR will be shown. In the second part of the seminar, classic TomoSAR algorithms (Fourier and Capon Beamforming) together with a new proposed algorithm based on compressive sensing will be analysed under different forest and tomographic configurations. Finally, the importance of the tomographic configuration will be stressed, first by showing results from single-pass (like Tandem-L) and repeat-pas (like Biomass mission) configurations and second by using configurations with different number and distribution of baselines.
Hannes Kistner „Antenna Pattern Correction in Elevation for New SIR-C processor“
Dienstag, 14. Mai 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
The Spaceborne Imaging Radar-C (SIR-C) was a radar system, which flew on two separate Space Shuttle missions in April and October 1994. Throughout these missions, a total of 143 hours of SAR data were recorded in multiple frequencies and polarizations. Unfortunately, high resolution SIR-C products are no longer available as the original SIR-C processor is no longer functional. To recover this historic dataset, the Alaska Satellite Facility is currently developing new software to process binary SIR-C data to high resolution, geocoded and calibrated SAR images. This bachelor thesis contributes to the relative radiometric calibration of the new processor by developing an approach to correct the antenna patterns in elevation. The main challenges of the SIR-C antenna pattern correction were the multitude of different antenna beams generated by active antenna steering and the parameter uncertainties in the new processor. Implementation and verification of an antenna model, which derives the antenna elevation diagrams required for accurate image correction, is therefore a key element in the presented work. The antenna model was implemented using probe data from pre-flight feed current measurements and verified with in-orbit antenna elevation pattern estimates derived from real SIR-C Amazon rain forest acquisitions. It was found, that the uncertainty of the available SIR-C ephemeris data is too high in order to determine the antenna pointing direction with sufficient accuracy. Hence, one of the main challenges was the development of an approach, that provides reliable estimates of the geometric parameters based on the image data itself. A two pronged approach was implemented based on in-orbit antenna pattern estimates on one hand and beam nulling on the other. Beam nulling is a calibration method employed by SIR-C, which involved inverting the receive phase of half of the antenna array in elevation by 180 degrees. Thus the effective beam received has a notch or null at the boresight. The final implementation of the antenna pattern correction algorithm was tested on a total of 12 real SIR-C datatakes, which equates to about one terabyte of raw data. The test results were found to be consistent with original Jet Propulsion Laboratory technical reports and show that the antenna model approach was successfully applied. In-orbit measurements of the residual relative radiometric cross-swath errors after antenna pattern correction were better than ±1 dB, with best case scenarios as low as ±0:11 dB over the Amazon.
Georg Fischer "Towards the Retrieval of InSAR Penetration Bias and Vertical Subsurface Structure of Ice Sheets by Means of Pol-InSAR Data"
Dienstag, 16. April 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
The penetration of microwave signals into snow and ice introduces a bias in digital elevation models generated by means of synthetic aperture radar (SAR) interferometry, particularly at longer wavelengths. This bias depends directly on the vertical backscattering distribution in the subsurface. In addition, polarimetric SAR interferometry (Pol-InSAR) and SAR tomography techniques allow investigating the vertical distribution of scattering processes, and have therefore the potential to provide geophysical information about the subsurface structure of glaciers and ice sheets. In this seminar, results based on F-SAR data from the ARCTIC15 campaign in Greenland are presented. Vertical backscattering distributions are investigated on two test sites in the percolation zone of Greenland using fully polarimetric X-, C-, L-, and P-band data. SAR tomography shows clear differences in the subsurface structure of the two test sites that belong to the same glacier zone. Based on the findings that existing models are insufficient to interpret the data, improvements in the interferometric modelling of the vertical backscattering distribution will be presented, which will support subsurface structure retrieval and penetration bias estimation. The assessed models improve the representation of the data compared to existing models while the complexity is still low to enable potential model inversion approaches. The tomographic analysis and the model assessment is therefore a step forward towards subsurface structure information and penetration bias estimation from SAR data. The seminar will conclude with an outlook on the retrieval of the penetration bias in InSAR DEMs.
Andreas Benedikter "Analysis of a Side-Looking SAR for Subsurface Imaging of the Ice-Moon Enceladus"
Freitag, 12. April 2019
10.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
SAR remote sensing may offer a key element for the exploration of Saturn’s ice-covered moon Enceladus, as it allows for imaging of the Enceladean ice-crust. As a consequence of the penetrating capabilities of electromagnetic waves up to a few gigahertz, the subsurface imaging of ice structures becomes feasible. In this work, the imaging characteristics of an orbital SAR mission over Enceladus are investigated. The main factors influencing the SAR imaging properties for such a mission are the unique orbit geometries and the dielectric properties of the Enceladean ice-crust. The proximity of Saturn combined with its huge mass have a strong influence on the orbit mechanics constraining the possible orbits for an Enceladean SAR orbiter, which are expected to for very long integration times in the range of 1000s to 3000s in low frequency ranges, which offers the potential of providing resolution in 3-D with a single image. Simulation results, backed by analytical expressions, show the imaging characteristics by means of the 3D impulse response of the system. The knowledge of the imaging characteristics is used to state the problem of accurate subsurface imaging for the case in which the ice permittivity is not a priori known. A multi-pass approach exploiting the defocussing of the scene is suggested for the accurate imaging of the Enceladean ice-crust.
Felipe Queiroz de Almeida "Slow PRI Variation: mode description and discussion"
Dienstag, 2. April 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
The purpose of the Seminar is to describe the rationale and basics of the Slow PRI Variation Mode, which is a new acquisition mode aimed at high-resolution wide swath applications. Employed in combination with multiple elevation beams, the mode has the peculiarity of a slow PRI variation (around 4% of the mean PRI) which leads to a smooth migration of the gaps due to Tx events in a wide swath configuration. Unlike Staggered SAR (fast PRI variation), the gaps are too large to be interpolated across, and lead to Doppler spectral gaps which change their position with slow time. The mode thus retains some properties of a “burst mode” like ScanSAR, though the illumination is more efficient than the former (meaning a finer resolution is possible) and the width of the gaps can be adjusted by changing the pulse duty cycle. The SAR signal properties and some of the identified advantages (e.g. shorter antenna) and disadvantages (e.g. higher power consumption that a Stripmap or Staggered-SAR with the same resolution) will be discussed. Some study cases in which the mode was found to be useful will also be described.
Junjun Yin "A unified framework for compact polarimetric SAR imagery"
Mittwoch, 27. März 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
Compact polarimetry (CP) is a dual-polarization hybrid imaging mode and provides partially backscatter characteristics of scatterers. The compact polarimetric synthetic aperture radar (SAR) measures combinations of fully polarimetric scattering coefficients, which is dependent on the transmit wave polarization state. The backscattered wave is fully described by a 2-dimensional complex scattering vector, in which the dual-channel polarization ratio is important for describing target scattering properties. However, due to the dependency of compact measurements on the transmit wave, the complex channel ratio is explained differently for same scatterers under different observation modes, e.g., the channel ratio ρ=±j under the left circular mode and ρ=±1 under the linear /4 mode for the canonical trihedral and dihedral scatterers. The explanation diversity is inconvenient for the use of compact data and results in non-unified target decomposition algorithms. A new formalism method is proposed for the first time for the general compact SAR descriptor, such that the compact measurements are described based on a same standard for target scattering representation. Then, based on this formalism, a polarization ratio-based scattering characterization method is proposed, which is mathematically equal to the fully polarimetric Δα_B/α_B method but without any scattering assumption. The formalism provides a natural way to characterize the target scattering behaviors. Experiments show that the proposed method is efficient for compact polarimetric scattering interpretation.
Diego Lorente Catalan "L-Band Antenna Array for Next Generation DLR Airborne SAR Sensor"
Dienstag, 19. März 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
The first concept of the new airborne L-Band antenna for the next F-SAR generation will be presented in this seminar. Due to the interesting and demanding applications of SAR at low frequencies, a new dual polarized airborne antenna system working in L-Band is required. Despite the always restricted available area in the airplane for the antenna installation, the new antenna system maximizes the number of elements by providing up to 66% more elements, in comparison with the current generation sensor. This improvement enhances significantly the radiation properties of the antenna system by allowing us to steer more accurately the radiation pattern. A special effort has to be made regarding the antenna compaction in order to exploit the most the limited available space for the antenna. In this seminar, the design of the single element antenna will be validated by means of comparing the simulation results with the measurement of a real manufactured antenna. Furthermore, the single antenna design is extended and used to build a linear array of five identical elements with amplitude and phase distribution in order to shape the radiation pattern as required. This antenna array is also manufactured and measured. The conclusion and future work will also be discussed. This new antenna concept is not only relevant for the next generation of our airborne SAR sensor but also for upcoming projects such as DuoLIM.
Ronny Hänsch "Feature Learning for Semantic Segmentation of PolSAR Images"
Donnerstag, 14. März 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
The one dominating factor for the current success of modern machine learning methods in computer vision and remote sensing are neither larger datasets nor stronger computational power but the creation of more descriptive features. While traditional approaches rely on domain-specific expert knowledge to carefully hand-craft and manually select features for a specific task (e.g. the mapping of PolSAR data to semantic labels), modern approaches include the computation of features into the optimization process of the machine learning framework itself. This renders the manual design of features obsolete and allows an end-to-end training of a single system that directly connects input and target variables. This talk presents two approaches of feature learning in the context of semantic classification of PolSAR images: A shallow approach based on a specific Random Forest variant as well as a deep learning based approach based on generative adversarial networks. It shows that despite the huge potential of and large success of deep networks, shallow methods are competitive in certain application scenarios and can offer benefits where deep models still have limitations.
Stefan Seel "Enabling higher data rates and Science Return in Earth Observation missions – a developments update"
Dienstag, 12. März 2019
14.00 h Großer Besprechungsraum HR, Gebäude 102
Abstract:
Earth Observation Systems with high resolution are designed for fast revisit time and for decreasing latency of data. To meet these demands, higher rates for data repatriation via direct downlink or Geo-relay are needed. ESA/EU Copernicus Sentinel satellites (1a/1b: C-band Radar, 2a/2b multispectral imaging) today significantly exceed their data downlink performance goals. Routine operations of laser relay to Geo-stationary EDRS-A accumulated more than 20,000 optical links. Mission benefits are in data quantity (50%), flexibility and in quality. Sentinel-1A can now operate in dual-polarization mode over main interest Europe land mass. Long term monitoring has confirmed that Key Performance Indicators (link quality and availability) are stable and better than currently exploited. What is next? The next generation LEO user terminal (SMART-LEO) will transfer the excess link margin into LEO user benefits (smaller and less power consumption) for links up to 1.8 Gbps. The next generation GEO terminal (NGGL-GEO) will be able to process at least double the data rate (3.6 Gbps). The presentation will report on the status of the developments and the upcoming in-orbit verifications (Pleiades Neo & Globenet-D). Of special interest to TESAT is feedback from the future user community with respect to laser terminal performance & service expectations. Complementary to optical solutions, the conventional direct-to-Earth RF-downlink payloads are seeing significant improvements. Adaptive Coding & Modulation (ACM) allows a much more efficient usage of available X-Band and Ka-Band bandwidths. At the same time, high performance forward error correction coding and high order modulation schemes are used. Results of a recent end-to-end performance verification of a prototype transmitter and ground station demodulator will be presented. To visualize the impact on payoad level, a highly integrated transmitter package suited for SmallSats will be presented. Potential benefits to Radar Sensor Solutions will be adressed in an overivew of the upcoming High Density Integrated technology developments at TESAT. The presentation will show the current development status of the pulsed Transmit/Receive Modules in X and L-Band and the Radar Frequency Equipment, using a High Density Integrated RF and Amplifier Module Technology.
Copyright © 2024 Deutsches Zentrum für Luft- und Raumfahrt e.V. (DLR). Alle Rechte vorbehalten.