Urban environments still form a challenge for Global Navigation Satellite System (GNSS) positioning due to its diffcult conditions for signal propagation. Many obstruction sources disturb the GNSS signal, signal loss or multipath are examples for effects occurring in these cases.This work focuses on an approach which distinguishes visible line-of-sight (LOS) satellite signals from obstructed non-line-of-sight (NLOS) signals. These are determined in a self-developed algorithm with the help of a 3D CityModel that is provided by the city of Hanover. For the evaluation, a kinematic experiment is conducted with geodetic and high-sensitivity receivers in a repeated trajectory. Further sensors were used to generate a reference trajectory.Open questions not only concern the signal characterization with respect to LOS/NLOS properties but also to what extent NLOS observations can be used in a constructive sense. Therefore, reflection points of NLOS signals are calculated, examined and introduced in a Single Point Positioning (SPP) solution.Additionally, these issues are put in the context of different receivers, each of which has different properties and hence diverse outcomes.Results show that the LOS status can be determined reliably but that the computed detour lengths not always match the Observed Minus Computed (OMC) values. Taking only those additional path lengths into account that fit the OMC, an improvement could be achieved in an SPP solution. Further findings show that OMC are especially high in street sections that have a perpendicular orientation compared to the satellite's azimuth angle.