I. Introduction
Absolute and relative positioning of vehicles is required to support a variety of autonomous drive (AD) and advanced driver assistance system (ADAS) services. While accurate relative positioning is enabled with sensors such as radar, camera, lidar [1], and inertial measurement unit (IMU), absolute positioning relies on an external reference. Such a reference can be in the form of a global map (so that local features can be associated in the global map), or in the form of satellites in a global navigation satellite system (GNSS) [2]. When maps are erroneous or out-of-date [3] or insufficiently many satellites are in view, global positioning is compromised [4], leading to safety risks or driver comfort degradation. Hence, alternative absolute positioning sensors, complementing high-resolution maps and GNSS are of high importance.