A Car2X Sensor Model for Virtual Development of Automated Driving

Automated driving requires a reliable digital representation of the environment, which is achieved by various vehicle
sensors. Wireless devices for communication between vehicles and infrastructure (Car2X communication) provide
additional data beyond the vehicle’s sensor range. In order to reduce the amount of on-road testing, there has been
an increased use of numerical simulation in the development of automated driving functions, which demands accurate
simulation models for the sensors involved.
The present research deals with the development of Car2X sensor models for conceptual, automated driving investigations
based on relatively simple yet computationally efficient mathematical models featuring parameters derived from on-road
hardware testing. For analysis purposes, variations in range and reliability in different driving situations were measured
and depicted in GoogleEarth. For the sensor model, a combination of geometric and stochastic models was chosen. The
modeling is based on a link budget calculation that considers system and path losses, where wave propagation is described
using Nakagami probability density functions. For intersections, an additional term is added to account for the path loss
with geometric parameters of the intersection. After model parametrization, an evaluation was conducted.
In addition, as a sample case, Car2X was added to an Adaptive Cruise Control (ACC), and the improved functionality was
demonstrated using vehicle dynamics simulation. This extended ACC used information from the indicator of surrounding
vehicles to react faster to lane changes by these vehicles.