Resource and Location Optimization with GA in UAV-Supported Multicellular Networks: A Comparison of Greedy and Local Search

Abstract

In this study, subband assignment to base stations (BS) and simultaneous optimization of the three-dimensional positions of unmanned aerial vehicles (UAVs) in a multi-cell network supported by UAVs were performed using a genetic algorithm (GA). The aim of the optimization is to maximize a total speed metric that prioritizes edge users (UEs) (based on max-min fairness) at the cell edge. The proposed GA was compared with improved greedy and greedy+local lookup methods. Simulations for different cell radii show that the GA outperforms both methods in terms of total speed and max-min fairness; it improves service quality, especially in large cells, by more effectively directing edge users to UAVs.