PROJECT TITLE :

Millimeter-Wave Base Stations in the Sky: An Experimental Study of UAV-to-Ground Communications

ABSTRACT:

In this paper, a systems approach is taken to investigate how millimeter wave (mmWave) radio transmitters on unmanned aerial vehicles (UAVs) provide high throughput links under conditions that are typical of hovering. On a testbed consisting of DJI M600 unmanned aerial vehicles (UAVs), we conduct experiments using Terragraph channel sounder units to investigate the impact of signal fluctuations and sub-optimal beam selection. We develop and validate the first stochastic UAV-to-Ground mmWave channel model using UAVs as transmitters. These developments are based on the insights related to hovering, as well as the measured antenna radiation patterns. Our analytical model that is centered on UAVs supplements the traditional fading with additional losses that are anticipated in the mmWave channel while hovering. This model takes into consideration the 3-D antenna configuration and beamforming training parameters. We pay particular attention to lateral displacement, roll, pitch, and yaw, the magnitudes of which change depending on whether or not specialized hardware such as real-time kinematic GPS is available. Then, we make use of this model to select a pair of beams that is nearly optimal in order to reduce the impact that hovering has on the connection between the UAV and the ground. Importantly, our work does not change the wireless standard in any way, nor does it require any cross-layer information. This ensures that it is compatible with mmWave devices that are currently in use. Results show that our channel model reduces estimation error to 0.2 percent, which is 18 times lower, and improves the average PHY bit-rate by 10 percent when compared to existing state-of-the-art channel models and beamforming methods for UAVs. This is demonstrated by a significant reduction in the standard deviation of the estimation error.