On Multi-Sensor Radar Configurations for Vehicle Tracking in Autonomous Driving Environments
Amanda Metzner, Thanuka L. Wickramarathne · 2018
Multi-sensor radar tracking systems can provide better system performance, visibility and complementary information than single sensor tracking systems. One way to utilize multiple sensors for target tracking is to generate individual sensor detections and fuse these detections to produce optimized system tracks. All or a subset of detections from the sensors are fed into a tracker algorithm, allowing the filter to choose the best possible measurement association. This paper discusses the implementation of a multi-sensor target tracking system using three different sensor configurations for providing vehicle detections. The three sensor configurations compare the accuracy of one vs two radar sensors, operating in either a stepped frequency waveform or a hybrid stepped frequency and continuous waveform. The purpose of this paper is to provide insight into the possibilities of multi-sensor radar tracking for the advancement of autonomous driving.