Single-frequency GNSS positioning for assisted, cooperative and autonomous driving

Abstract

Through the advent of more and extended GNSS constellations, Single Frequency (SF) positioning gains much in performance and applicability. In this contribution we explore the role SF GNSS can play for driver assistance, cooperative driving and fully autonomous vehicles. A convoy experiment was conducted with three vehicles, driving behind each other on a straight road. Each of the three vehicles was equipped with a SF GNSS receiver (u-blox M8T) and a patch antenna, and at the side of the road a fourth identical SF GNSS receiver with a patch antenna was setup. Raw data logged with these receivers were processed in several modes, exploring their possible use for different types and levels of vehicle automation. SF-PPP is considered for driver assistance, moving base (MB) differential processing (i.e. combined processing of the data from two or more moving vehicles) is considered in the light of cooperative driving, and RTK is considered for self-driving vehicles. In each of these modes, the data were processed as if it were in real-time. Reference data were collected with a professional robotized total station for land-surveying (Leica), setup at the road side independently measuring absolute vehicle positions with centimeter precision. Additionally, the second vehicle was equipped with a laser disto-meter (Leica) to independently measure the distance between the first two vehicles, with millimeter-precision. Results show that the accuracy requirements are met for each of the three applications considered. Availability is also high, and the advantages of a multi-constellation approach are highlighted. The results also contained evidence of a well-known weakness of GNSS, e.g. its sensitivity to RFI due to the low received signal power.