Trajectory Design for the Aerial Base Stations to Improve Cellular Network Performance

Aerial base stations (ABSs) have shown significant potential to improve cellular networks’ performance due to their high mobility and on-demand deployment. In this paper, the trajectory optimization problem for a multi-ABS network is investigated. The objective is to maximize the minimum data rate of the cell-edge mobile users while the power constraint of the ABSs, including both propulsion and signal transmission powers, the backhaul link capacity constraint, and the collision avoidance constraint, are taken into account. To reach this goal, first, based on the modified K-means approach, the ABSs and users are partitioned into different clusters so that their associated ABS serve the users of each cluster. Afterward, the ABSs need to find an efficient trajectory optimization and resource allocation scheme to support the users. To solve this challenging problem, we convert the main trajectory optimization and resource allocation problem into three sub-problems: power allocation sub-problem, joint ABS-user association and sub-channel assignment sub-problem, and trajectory optimization sub-problem. Then, using the successive convex approximation approach, an efficient algorithm is proposed, which iteratively solves these sub-problems. Simulation results show that the proposed algorithm converges fast and improves the network's data rate while it satisfies all the required constraints.