Optimal 3D Placement for Integrated Access Backhauling in UAV-Assisted Wireless Networks Using Reinforcement Learning

Unmanned aerial vehicles (UAVs) have emerged as a promising solution to augment the performance of existing cellular networks, especially for mission-critical operations. Integrated access and backhaul (IAB) in 5G networks is a potential solution that can leverage UAVs to enhance the coverage in under-served regions. However, the dynamic user locations and the need for balancing backhaul and fronthaul connectivity for end-to-end performance can be challenging. This paper addresses these challenges by proposing a dynamic, real-time approach based on reinforcement learning for end-to-end performance optimization in a UAV-assisted cellular network. The proposed methodology considers both fronthaul and backhaul links to optimize the 3D placement of the UAV, while ensuring high data rates for ground users. Our results indicate that the proposed method effectively balances the tradeoff between backhaul and fronthaul links to achieve maximize the end user performance.