High Fidelity Physics Based Simulation for Automotive Radar and V2X Applications
Abstract: Autonomy and electrification have emerged as key drivers of innovation and growth in the automotive industry in recent years. Various advanced driver assistance systems (ADAS) have been developed as OEMs race towards a fully autonomous and zero emissions future. Automotive radar and vehicle-to-everything (V2V/V2X) technologies are two of the key technologies that enable vehicles to have situational awareness while also communicating with other vehicles and road infrastructure. Here, we present high fidelity physics-based electromagnetics simulations for V2V and automotive radar systems. Specifically, we will demonstrate the impact of vehicular obstructions, multipath effects, and non-line-of-sight (NLOS) conditions on 5.9 GHz designated short-range communication (DSRC) links. Automotive radar simulations of a synthetic, 128-channel, 77 GHz, multiple input multiple output (MIMO) sensor for angle of arrival determination will also be presented. We will also present a micro-Doppler based convolutional neural network (CNN) for classification of vulnerable road users (VRUs). Here, we use simulation to obtain thousands of spectrograms spread across 4 target classes (car, pedestrian, cyclist and dog). Using these spectrograms, we train a 5-layer CNN to classify targets with an accuracy greater than 97%.