Scenario-based tool chain for virtual verification and validation of automotive radar

• Applicants: Professor Dr.-Ing. Thomas Dallmann; Professor Dr.-Ing. Lutz Eckstein; Professor Dr. Matthias Hein
• Subject Area: Communication Technology and Networks, High-Frequency Technology and Photonic Systems, Signal Processing and Machine Learning for Information Technology
• Term: since 2022
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 503852364

The safety and efficiency of connected and automated driving functions and driver assistance systems (ADAS) relies on wireless sensor perception, in addition and complementary to wireless communications. Hence, reliable and efficient testing presents an extremely urgent research task. While assuring the safety of ADAS and automated driving functions with a statistical approach would require billions of test kilometres, there is no doubt that, in view of the exploding number of sensors and complexity of driving functions, a scenario-based test approach is required. Of special relevance are challenging scenarios that exhibit a high risk of failure of the underlying intended function. However, until today, there is a lack of efficient testing methods and meaningful fidelity metrics that deliver reliable pass/fail criteria. Consequently, the proposed project focuses on the virtual verification and validation of automotive radar in a holistic multi-layer approach, reaching from software-in-the-loop to field tests on proving grounds, which connects seamlessly to existing generic approaches of scenario-based testing. The research is driven by the following research questions: 1. What are the limiting factors of scenario-based virtual verification and validation of auto-motive radar and how can they be overcome by a holistic approach? 2. How can relevant scenarios be described and realised in the virtual domain? 3. How can test metrics and figures-of-merit for wireless advanced driving functions be de-fined and evaluated? With the expertise of Prof. Eckstein, precise reference data will be collected by novel drone-based techniques, to identify challenging scenarios from a measurement point-of-view. This task is completed from a propagation and environmental modelling point-of-view with the expertise of Dr. Dallmann and Prof. Hein (Objective 1). The anticipated results will lead to improve radar perception algorithms, extend the modelling by clutter and interference, and substantiate the emulation of scenarios in virtual environments with the over-the-air vehicle-in-the-loop approach (Objective 2). The third objective deals with the composition of a test catalogue and recommendations on how to test wireless automotive functionalities like radar and communication, based on test metrics for the sensors themselves as well as the test method as a whole. In addition, cross-metric figures-of-merit will be derived, by combining the various sets of metrics, in order to evaluate the quality of advanced driver assistance systems and their functionality on a reproducible level. This project forms an integral package together with the project P1. The two projects are tightly tied through advanced dynamic ray tracing methods as well as cross-metric quality evaluation of the performance of the scenario-based tool chain for virtual verification and validation.

• DFG Programme: Research Grants
• Co-Investigator: Dr.-Ing. Christian Bornkessel