Ultra Wideband Radar for Environmental Polarimetric Observation and Scattering Analysis

• Applicant: Professor Dr.-Ing. Reiner Thomä
• Subject Area: Electronic Semiconductors, Components and Circuits, Integrated Systems, Sensor Technology, Theoretical Electrical Engineering
• Term: from 2014 to 2018
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 255934852

The Ultra EPOS project addresses polarimetry in ultra wideband (UWB) radar applications. UWB radars are well known for their excellent time resolution in short range localization, tracking and imaging applications. However, exploitation of UWB polarimetric information is still in its infancy despite the fact that the polarimetry is renowned in the field of narrowband SAR remote sensing (earth observation satellites, airborne imaging systems). The main goal of this project is to fill in this gap and to bring advantages of polarimetry into the field of UWB radars as well. UWB polarimetry is anticipated to provide more detailed information about targets and structures such as shape, orientation (edges, corners), material composition or surface roughness. In contrast to the classical remote sensing methods which analyse polarimetric information in the image (focused) domain, UWB polarimetric signal processing may already be applied in the measurement (time) domain. The large bandwidth of UWB sensors allows temporal separation of measured polarimetric signal components. Therefore, signal features can be assigned to a specific scattering process already in the time domain. This will significantly influence imaging techniques, object detection methods and polarimetric filters that can amplify or suppress specific signal components. Polarimetric processing will enable improvement of clutter filtering and support exploitation of multipath reflections from the environment to enhance scene illumination. In our assumed scenarios, transmitters and receivers are in a multistatic arrangement. They can be moved along arbitrary tracks in order to gain spatial diversity by illuminating and observing targets and environmental structures from different multistatic angles. The resulting geometric constellation is completely different from what we know in classical remote sensing and requires new methods for exploitation of UWB polarimetric information in imaging and detection. The new UWB polarimetric imaging algorithms will be based on models that represent different polarimetric fingerprints. This allows the construction of distinct images form the respective object model classes. In this way clutter will be reduced. Data fusion of multiple partial images will lead to a clearer overall image of the target object and allow more reliable interpretation on target shape, material composition and surface structure. We consider two complementary scenarios. The first scenario encompasses solitary target objects. We perform extensive measurements of selected objects in order to obtain a library of basic polarimetric fingerprints that are not influenced by a disturbing clutter due to the environment (walls, other objects). In a second realistic indoor scenario this target-to-environment interaction shall be exploited to obtain more information about the target (multipath exploitation) and to improve detection and imaging performance.

• DFG Programme: Research Grants