Roadside Sensor Optimization in Infrastructure-based Teleoperation of Vehicles

Inhalt:
Infrastructure-based teleoperation relies on roadside sensors and offers an alternative approach for data transmission in teleoperation of remote vehicles. Since the number of deployed sensors within a field of interest is often limited by physical or economical constraints, their placement is of fundamental importance. This paper describes a self-learning solution for the sensor placement problem based on Fritzke’s Growing Neural Gas algorithm. The proposed method iteratively computes the sensor locations, trying to reach a given road coverage threshold.