Sie sind hier: Home:Abteilungen:Aufklärung und Sicherheit

Abteilung Aufklärung und Sicherheit (Dr. Thomas Neff)

Die Abteilung Aufklärung und Sicherheit bearbeitet Themen aus dem Bereich Fernerkundung und Aufklärung für hoheitliche, nationale Systeme. Das Hauptaugenmerk der Arbeiten liegt dabei auf:

Aufklärungssysteme

Entwicklung und Realisierung der notwendigen Entwurfs-, Analyse- und Simulationswerkzeuge für raumgestützte SAR-Systeme mit höchster räumlicher und radiometrischer Auflösung einschließlich der notwendigen Signalverarbeitungsverfahren,
Analyse des Rückstreuverhaltens von natürlichen und von Meta-Materialien,
Erarbeitung von Methoden und Werkzeugen zur effizienten Informationsgewinnung aus Bilddaten,
Aufbau von bodengebundenen SAR-Experimentalsystemen mit höchster räumlicher und radiometrischer Auflösung,
Aufbau von vollpolarimetrischen Mikrowellenradiometern nach dem Prinzip der Apertursynthese.

Die Abteilung ist in fünf Fachgruppen unterteilt, deren Ziele und Themen im Folgenden beschrieben werden.

Forschungs- und Projektgruppen

Prof. Lorenzo Bruzzone "Challenges of Remote Sensing Data Analysis in the Artificial Intelligence Era"

Freitag, 7. Februar 2020
10.00 h Großer Besprechungsraum HR, Gebäude 102

Abstract:
This talk addresses the status and the challenges of remote sensing data analysis in the artificial intelligence era, where deep learning is becoming the dominating processing paradigm. Despite the large success of deep learning in computer vision, where deep architectures provided groundbreaking results in solving operational problems related to classification and semantic segmentation of big data, the use of deep learning in operational remote sensing still requires addressing many crucial challenges. On the one hand, there is the need to define architectures that can properly model the properties of different kinds of remote sensing data (e.g., multispectral, hyperspectral, SAR, LiDAR). On the other hand, there is the intrinsic limitation on the available reliable labeled data required for a robust training of the deep architectures in a scenario where different applications often require the recognition different target classes according to different legends. In the talk, these challenges are presented and critically analyzed, identifying high level strategies for addressing them. This involves the analysis of learning components, learning paradigms, available data sources for the definition of large training sets, integration with physical models as well as exploitation of advanced methodologies related to active learning and domain adaptation. This will be discussed presenting hierarchical strategies for the definition of a common approach to the development of sensor specific and application specific processing architectures.

Eduardo Rodrigues Silva Filho "GNSS-based Position and Baseline Determination and Clock Synchronization of Multistatic SAR"

Donnerstag, 30. Januar 2020
14.00 h Großer Besprechungsraum HR, Gebäude 102

Abstract:
Multistatic constellations can offer various advantages for SAR remote sensing. These concepts are challenging to implement for a series of technical difficulties. The lack of synchronization, caused by the operation of transmitter and receiver with different master clocks, poses one of the fundamental operational problems, contaminating the phase signatures of the radar imaging and challenging its differential ranging accuracy. In addition, baseline accuracy of a few millimeters must achieved, preferably using data obtained from low-cost GNSS receivers. In this work, we evaluate a synchronization method based on GNSS navigation data and Precise-Orbit Determination. The method consists in using in each satellite the same oscillator for the master clock of the GNSS receiver and of the SAR payload, so that the relative time estimation obtained in the precise orbit determination can be used to synchronize the radar data in the post-processing. The simulations suggest the proposed approach is capable of delivering reliable estimates of phase errors in the absence of strong baseline velocity deviations and if multipath and other systematic errors are successfully suppressed or calibrated. In addition, we evaluate different configurations in an attempt to improve the individual baselines estimates by combining GNSS data from several satellites flying in close formation. In our preliminary studies we conclude that the individual baseline can potentially be improved by using intersatellite links and by implementing a consistency check by comparing the height biases between DEMs generated from different pairs of satellite.

Koenraad Mouthaan „Modeling of reflector based antennas and digital beamforming using Keysight's Advanced Design System (ADS)”

Dienstag, 28. Juli 2020 um 14.00 h
Dieser Vortrag wird virtuell stattfinden! - Die Log-in Daten werden zeitnah bekanntgegeben.

Abstract:
With the ever increasing demands on new SAR systems, accurate modeling becomes more important in the early stages of the design of the systems. For example, in reflector based systems the mutual coupling within the feed array and with the reflector, as well as non-ideal frequency dependent behaviors, impact the performance of the system. Having an accurate model available supports the work of system architects and system designers. Keysight's Advanced Design System (ADS) is a software tool for the design and analysis of microwave and high speed components, circuits, and (sub-)systems. ADS supports various types of analysis such as DC, AC, small-signal scattering parameters, large-signal simulation, and time-domain. This presentation explores the use of ADS for the modeling of reflector based antennas and digital beamforming. The modeling allows the incorporation of non-ideal behaviors through simulated data, obtained from electromagnetic simulators, or measured data. The presentation starts with a short introduction of ADS and the motivation, followed by a brief introduction of scattering parameters. Subsequently the modeling of antennas, using electromagnetic data generated by ANSYS HFSS, is demonstrated. Several short case studies, including a planar array and a reflector based antenna, illustrate the modeling approach. Finally, conclusions and recommendations are presented.

Paul Kroll „Internal Instrument Calibration – Mathematical Models and Simulations“

Dienstag, 14. Juli 2020 um 14.00 h
Dieser Vortrag wird virtuell stattfinden! - Die Log-in Daten werden zeitnah bekanntgegeben.

Abstract:
Two key requirements for a modern, space-borne Synthetic Aperture Radar are a high azimuth resolution and a wide swath. This can be achieved by the use of a SAR instruments with multiple digital channels. Differences between the channels reduce the performance due to the degradation of the antenna pattern. To compensate the differences the system has to be calibrated. To describe the behavior of such a system, mathematical models were developed, based on realistic hardware behavior. These models are used to represent the signals, the instrument drift, the temperature distribution, error sources and other physical effects. Further, a simulation tool, to estimate the errors, residual errors and aid in the system design, was implemented. The tool is adaptable to any multi-channel system and can be used for different calibration methods (e.g. single/multiple calibration signals, chirp signals, etc.). An internal calibration concept/methodology is suggested which allows for a simultaneous calibration during SAR transmit and receive operation. For the receive calibration, a single tone calibration signal, the CalTone, is sequentially coupled to the echo signal path. The transmit calibration is done by first characterizing the receiver (radio frequency unit and digital beam-forming unit) with a CalTone and then couple a part of the transmitted signal to the receiver. To account for frequency dependencies the frequency of the calibration signal can be varied. This presentation will show the above mentioned models, concepts and simulation results.

SAR-Analysen (Dr. Björn A. Dietrich)

Die Fachgruppe SAR-Analysen entwickelt analytische und numerische Verfahren für den Entwurf und die Analyse von raumgestützten SAR-Systemen für Aufklärungsmissionen.

Mehr

SAR Simulation (Dr. Rainer Speck)

Kontakte

Dr.-Ing. Thomas Neff
Abteilungsleitung: Aufklärung und Sicherheit
Deutsches Zentrum für Luft- und Raumfahrt (DLR)
Institut für Hochfrequenztechnik und Radarsysteme, Aufklärung und Sicherheit
Oberpfaffenhofen-Wessling
Tel.: +49 8153 28-3327
Fax: +49 8153 28-1135

Dr.-Ing. Björn A. Dietrich
Gruppenleitung: SAR-Analysen
Deutsches Zentrum für Luft- und Raumfahrt (DLR)
Institut für Hochfrequenztechnik und Radarsysteme, Aufklärung und Sicherheit
Oberpfaffenhofen-Wessling
Tel.: +49 8153 28-3073
Fax: +49 8153 28-1135

Dr.-Ing. Rainer Speck
Gruppenleitung: SAR-Simulation
Deutsches Zentrum für Luft- und Raumfahrt (DLR)
Institut für Hochfrequenztechnik und Radarsysteme, Aufklärung und Sicherheit
Oberpfaffenhofen-Wessling
Tel.: +49 8153 28-2322
Fax: +49 8153 28-1135

Dr.-Ing. Markus Peichl
Gruppenleitung: Mikrowellensensorik
Deutsches Zentrum für Luft- und Raumfahrt (DLR)
Institut für Hochfrequenztechnik und Radarsysteme, Aufklärung und Sicherheit
Oberpfaffenhofen-Wessling
Tel.: +49 8153 28-2390
Fax: +49 8153 28-1135

Dr.-Ing. Jan Eilers
Gruppenleitung: Systemtechnik
Deutsches Zentrum für Luft- und Raumfahrt (DLR)
Institut für Hochfrequenztechnik und Radarsysteme, Aufklärung und Sicherheit
Oberpfaffenhofen-Wessling
Tel.: +49 8153 28-3980
Fax: +49 8153 28-1452

Dr. tech. Erich Kemptner
Gruppenleitung: Signaturen
Deutsches Zentrum für Luft- und Raumfahrt (DLR)
Institut für Hochfrequenztechnik und Radarsysteme, Aufklärung und Sicherheit
Oberpfaffenhofen-Wessling
Tel.: +49 8153 28-2370
Fax: +49 8153 28-1135