Interference measurements between single-beam, mechanical scanning, time-of-flight lidars

Abstract

Signal interference between light detection and ranging (lidar) sensors has the potential to degrade range data integrity, which may challenge multilidar applications such as autonomous vehicles if not adequately considered. Our work proposes a methodology for evaluating lidar interference, analyzes distinct lidar interference phenomena, and validates a previously proposed model of interference between two circularly scanning, pulsed lidar sensors. The comparison between Monte Carlo simulated and experimentally observed interference events suggests that lidar interference may be inferred through geometrical approximations of lidar scanner arrangements. The experimental data quantify the occurrences of at least two modes of lidar interference—direct and scattered interference. When present, direct interference was found to occur one to two orders of magnitude greater in total occurrence than scattered interference. However, direct interference’s effects represented an average range error of <1 m. Alternatively, scattered interference was observed with an average range error two to four orders of magnitude greater than direct interference and proportional to the maximum range of the victim lidar. Further characterizations of interference phenomenon are presented that include the angular distribution of erroneous ranging data, range distributions of errors, loss of in-tolerance points, and potential radiometric influences.