Theoretical Analysis and Active Control for UAV Millimeter Wave Communications
Abstract
The unmanned aerial vehicle (UAV) communications have been an important part of the space-air-ground integrated networks, thanks to the flexible coverage and line-of-sight (LoS) propagation. But the goal of high-speed data transmission relies on efficient alignment of narrow high-gain beams between the UAV and the base station (BS), which is often highly challenging, especially considering UAV mobility and millimeter wave scenarios. To tackle this challenging issue, the idea of control-assisted communications was recently proposed, where UAV-control signalling is exploited for efficient beam prediction and tracking. The developed approaches to date are heuristic, since no analytical framework exists to characterise the dependence of UAV beam direction variation on the control system parameters. In this paper, we fill this gap by first deriving a model for beam direction dynamics, with control signal as model input. Then, we reveal theoretically how parameters like proportional gain of the control system affect the performance of beam tracking, by studying the rise time and maximum overshoot metrics of the controller. We also analyze and reveal the important tradeoff between performance metrics for communication and control. With the beam direction dynamics available, we further propose optimizing communication system via active control. The core is to design or optimize the control signal to adjust UAV flight and thus change the communication channel, so as to achieve better communication performance or desired tradeoff between the communication and control. Finally, we design a simple but effective input control signal. Simulation results demonstrate that with the help of our active control approach, even a simple beam tracking algorithm can achieve better performance.