Radar image x and y with two targetsInterfered time domain signalCar as target with mutliple reflection centers and box model

Radar Signal Processing, Modulation, and Interference

As part of its research in the field of automotive radar, UAV, and industrial applications, the Institute of Microwave Engineering works on improving and enhancing the underlying signal processing methods as well as on developing new concepts and methods for current and future radar systems.

Together with various industry partners, adaptive modulation methods are being investigated in order to accommodate the ever changing demands of traffic situations. In the case of the well-established chirp-sequence modulation, the combination of multiple measurements, and the variation of the ramp slope allows the radar to flexibly adjust its parameters, like resolution, thereby adapting to the current traffic situation. New methods for signal generation and processing for future fully-adaptive digital radar approaches like OFDM and PN (pseudo noise) are being investigated and further developed.

Another research area at the Institute is that of radar interference. The increasing number of radar sensors in automotive scenarios increases the probability of interference between the sensors, leading to a reduction in sensor operability up to the total loss of operation. Impact and properties of this interference are being investigated and counter-measures developed, using approaches like frequency hopping or digital beamforming. Hereby, interference effects between FMCW-, chirp-sequence-, PN-, and OFDM-modulated radars are considered.

In addition to hardware related signal processing, higher level radar processing is an area of research for the Institute and its industry partners. The combination of next-gen high-resolution radars at 77 GHz and newly developed algorithms allows for the detection, distinction, and classification of extended objects like vehicles on the road based on their estimated length, width, and orientation. Using machine learning methods, motion and gestures of pedestrians can be recognized. These new methods allow future driver assistance systems and autonomous vehicles a more accurate and robust analysis of traffic scenarios and the early detection of potential dangers with a higher reliability than traditional techniques.

Furthermore, the Institute investigates complex synthetic aperture (SAR) processing methods, that allow the extraction of ultra-high-resolution radar images of the environment on the fly. For this, algorithms to compensate for Doppler- and range-migration effects due to the ego-motion of the sensor and other nonlinearities are being investigated and developed.

Skills / Expertise

  • Adaptive radars and modulation
    • OFDM radars
    • PN Radars
    • Chirp-sequence radars
  • Interference prevention and mitigation
  • Object detection and classification
  • Gesture recognition using machine learning
  • SAR processing

Funded Projects
In addition to a variety of bilateral industrial projects the following research projects are handled:

  • TC a-drive: "Adaptive automotive radar sensors"
  • KoRRund: "Radar sensor networks for automotive applications"
  • Radar4FAD: "Code modulated automotive radar sensors"
  • RobustSense: "Polarimetric automotive radar sensors"