Experimental Investigation of Pointing Errors on Drone-based FSO Systems

Airborne free space optical (FSO) links can be efficiently used for backhaul/fronthaul needs of future cellular networks. Unlike terrestrial FSO links where a fixed connection between two buildings is considered, airborne FSO links bring unique challenges due to the mobility of the drone. In this paper, we carry out an experimental study to demonstrate the effect of pointing errors in drone-based FSO systems. We employ a supervised learning method for tracking purpose of the FSO signal. We take power measurements and characterize the probability distribution of the received signal power. We further present comparisons between tracking and non-tracking cases and demonstrate the benefits of the tracking system.