Artificial Intelligence-Driven Unmanned Aerial Vehicle Base Station Placement: Current Advances, Challenges, and Use Case

In recent years, there has been a lot of interest in using drones as aerial base stations (BSs), namely drone base stations (DBS), to help with ground connection under unique situations. Despite the apparent benefits of DBSs, they present some significant obstacles that must be solved in order to provide a robust drone-assisted wireless network service. Optimal placement of DBSs in the air and their movement to ensure optimum coverage is one of the most crucial difficulties. In this paper, we provide an overview of DBSs and their challenges, focusing on the main challenge, the placement, and how artificial intelligence (AI) can be utilized to solve such a challenge effectively. Finally, we present a use case in which we propose an AI-based approach for distributing multiple DBSs to serve zones within a network while allowing for dynamic behavior in user demand and services over time. This problem is solved using the RL approach based on the SARSA algorithm for exploration and continuously learns and adapts DBS allocation strategies in response to environmental interactions. The simulation results show an increase in operator profit compared to equal policy dispatch.