Joint Trajectory and Frequency Offsets Design for FDA-Based UAV Communications

In this paper, we study how to improve the communication quality in millimeter-wave (mmWave) ground-to-unmanned aerial vehicle (UAV) communication systems. The system consists of a ground node with an antenna array and a UAV with a single antenna. Obviously, the communication quality can be maximally enhanced in the case that the trajectories of array beampattern peak and UAV coincide with each other. Achieving this objective through a traditional phased array (PA) approach requires considerable implementation costs, as the phase lags among array elements need to be frequently updated to change the beampattern peak direction. To alleviate this, we design an efficient method based on frequency diverse array (FDA), which tunes the frequency offsets among array elements. As a result, FDA generates a time-variant beampattern, such that the beampattern peak exhibits autonomous spatial movement. By using this characteristic, the complexity of the FDA method is reduced compared to PA, since we can design the beampattern peak trajectory without the necessity for real-time parameter updates. Specifically, subject to the constraints of UAV's speed, flight time and FDA's frequency offsets range, we jointly optimize the UAV movement and FDA frequency offsets to maximally align the trajectories of UAV and FDA beampattern peak, thereby improving beampattern gain. However, the formulated problem is non-convex with highly coupled variables. To handle this problem, we further design an algorithm based on penalty dual decomposition (PDD) method which circularly updates the variables, and finally yields a stationary solution of the problem. Simulation results demonstrate that, with the same implementation complexity, the FDA approach can achieve a higher communication quality compared to the PA approach and other benchmark schemes.