Performance Analysis for Autonomous Vehicle 5g-Assisted Positioning in GNSS-Challenged Environments

Abstract

Standalone Global Navigation Satellite Systems (GNSS) are known to provide a positioning accuracy of a few meters in open sky conditions. This accuracy can drop significantly when the line-of-sight (LOS) paths to some GNSS satellites are obstructed, e.g., in urban canyons or underground tunnels. To overcome this issue, the general approach is usually to augment GNSS systems with other dedicated subsystems to help cover the gaps arising from obscured LOS. Positioning in 5G has attracted some attention lately, mainly due to the possibility to provide cm-level accuracy using 5G signals and infrastructure, effectively imposing no additional cost. In this paper, we study the hybridization of GNSS and 5G positioning in terms of achievable position and velocity error bounds. We focus on scenarios where satellite visibility is constrained by the environment geometry, and where the GNSS and 5G positioning systems fail to perform individually or provide prohibitively large error.