Automotive Radar Signal and Interference Simulation for Testing Autonomous Driving

With the development of automated driving functions, more and more environmental sensors are combined for the vehicle perception. A problem that arises with the extensive use of radar sensing is called interference. It describes the confounding effects from the wave overlay of two or more radar sensors operating in the same frequency-band. At this point, methods for interference avoidance and mitigation come to apply. For a valid design and development of such methods, real sensor measurements were required in the past. This publication instead proposes a novel sensor modelling technique that represents the interference mechanisms within the radar sensor signals. It is based on a full radar time signal simulation coupled with a broad range of influencing factors. The concept is validated by comparing the simulated signal processing steps to the real sensor measurement behavior. As a result, mitigation methods for the sensor fault behavior can be fully assessed within a simulation environment. The opportunity for applying new scenario data and a variable set of radar sensors underlines the importance of this approach in the development of future radar systems.