Fixed Wing UAV-based Non-Terrestrial Networks for 5G millimeter wave Connected Vehicles

This paper analyzes aerial Radio Access Networks for enhanced mobility coverage due to superior Line of Site (LOS) access, which is vital for 5G and future 6G-connected at millimeter wave and higher bands. This paper analyzes aerial rotary-wing and fixed-wing unmanned aerial vehicle (UAV) access nodes, or gNBs, with millimeter-wave connectivity to ground vehicles. The scheme presented introduces generalized geometric and radio link simulation analysis of rotary and fixed-wing models. This paper also analyzes UAV-RAN access at altitudes of 20 Km, referred to as High Altitude Platform Stations (HAPS). The analysis presents 5G UAV-based gNB-RAN HAPS systems which support lower latency connectivity compared to Low Earth Orbit satellites, Furthermore, ultra-low latency capability supports tactile HAPS network control of autonomous ground vehicles. The received (Rx) data statistics and coverage establish a framework for Machine Learning (ML) based 2Dim multi-armed bandit optimization and models for 20 Km HAPS gNB systems operating as aerial millimeter wave access nodes. Simulation results demonstrate 5G connected vehicles' performance capability based on the multi-armed planar antenna array formed beams. Models employ a Butler network with array control of the amplitude and phase of antenna elements. Finally, metrics analysis for the Cumulative Distribution Function (CDF) of received data during the iterations of the machine learning algorithm illustrate excellent ground cell coverage. Cell radiuses of 4000 meters, 8000 meters, and 12000 meters at a gNB height of 20 km employ machine learning optimized beam selection based on a 2-Dim Multi-armed bandit model.