Measuring and optimizing for network conditions on drones

Abstract

As drones emerge as viable vehicles for mobile computation, a world where we can deploy a fleet of drones to execute tasks, like streaming real-time video, sounds less like sci-fi and more like a quickly-approaching reality. In order to take full advantage of the capability of drones, we must understand the state of network conditions in the air and determine how we can best optimize for them. Here, we contribute twofold. First, we developed and deployed a set of tools to collect network data on a flying drone and then process it to characterize the network conditions the drone experienced. Second, we developed a simulator to play back the collected network traces and help understand how to harness the power of a fleet of drones working in coordination to compensate for unstable network conditions in the air.