# Effects of 3D Antenna Radiation and Two-Hop Relaying on Optimal UAV Trajectory in Cellular Networks

## TL;DR

This study investigates how 3D antenna radiation and two-hop relaying influence the optimal trajectory of UAVs in cellular networks, optimizing for capacity and coverage under various constraints using dynamic programming.

## Contribution

It introduces a comprehensive trajectory optimization framework considering 3D antenna patterns, relaying, and mobility constraints, employing dynamic programming for UAV path planning.

## Key findings

- Maximum sum-rate trajectory yields highest per user capacity.
- Proportional fair trajectory improves coverage probability.
- Bezier curve smoothing maintains capacity close to optimal trajectories.

## Abstract

In this paper, considering an interference limited inband downlink cellular network, we study the effects of scheduling criteria, mobility constraints, path loss models, backhaul constraints, and 3D antenna radiation pattern on trajectory optimization problem of an unmanned aerial vehicle (UAV). In particular, we consider a UAV that is tasked to travel between two locations within a given amount of time (e.g., for delivery or surveillance purposes), and we consider that such a UAV can be used to improve cellular connectivity of mobile users by serving as a relay for the terrestrial network. As the optimization problem is hard to solve numerically, we explore the dynamic programming (DP) technique for finding the optimum UAV trajectory. We utilize capacity and coverage performance of the terrestrial network while studying all the effects of different techniques and phenomenon. Extensive simulations show that the maximum sum-rate trajectory provides the best per user capacity whereas, the optimal proportional fair (PF) rate trajectory provides higher coverage probability than the other two. Since, the generated trajectories are infeasible for the UAV to follow exactly as it can not take sharp turns due to kinematic constraints, we generate smooth trajectory using Bezier curves. Our results show that the cellular capacity using the Bezier curves is close to the capacity observed when using the optimal trajectories.