Probability of Line of Sight Evaluation in Urban Environments using 3D Simulator

The integration of Non-Terrestrial Networks (NTNs) into 6G networks is one of the most promising ways to achieve significant improvements in capacity, reliability, and global coverage. The design of NTN heavily relies on using channel models. In this paper, we propose two easy-to-use simulators for estimating the Line-of-Sight (LoS) probability $\mathbf{P_{LoS}}$ in a 3D urban environment. The first simulator is a 3D city simulator that employs simplified Ray Tracing (RT), while the second one is a lightweight geometry-based simulator that generates only the relevant buildings between users and Unmanned Aerial Vehicles (UAVs). Using these simulators, we assess the accuracy of existing models for $\mathbf{P_{LoS}}$ estimation and examine $\mathbf{P_{LoS}}$ for different UAV heights, user-UAV distances, and azimuth/elevation angles. We conclude that 1) existing models overestimate $\mathbf{P_{LoS}}$, resulting in overoptimistic NTN performance predictions, 2) nodes location (including azimuth and elevation angles) is an important factor influencing $\mathbf{P_{LoS}}$, however, this influence is not captured by the existing models.