Autonomous Formation Flying of Micro Aerial Vehicles for Communication Relay Chains

Kurzfassung

In this paper we describe the use of GNSS location determination for formation flying of multiple Micro Aerial Vehicles (MAVs). These MAVs are the core component of a system that realizes a chain of airborne communication relays to provide communication services to mobile terminals. This chain of communication relay nodes improves wireless radio communication by utilizing two effects. Firstly, a chain of relay nodes is able to convert an unfavorable non-line-of-sight condition to a line-of-sight condition by positioning its nodes around the obstacle. Secondly, relay nodes reduce free-space losses. The transmission of a radio signal in free-space is subject to a reduction in received signal power proportional to the square of the distance between transmitter and receiver. By placing a communication relay between the original transmitter and receiver, the distance is cut in half, reducing the required transmission power for each transmitter by a factor of four. The reductions in required transmission power can be directly utilized to increase the range and/or data rate of the communication system. The motors required to hover the relay platform also consume power, typically more than the transmission power. But as the transmission power of a single antenna is limited and cannot be increased over a certain level, range and bandwidth are limited. Flying relay nodes increase the necessary total electrical power, but offer the possibility to enhance range and data rate above the limits of a single node. Location determination of all involved network nodes, both airborne and on the ground, is crucial for determining and maintaining the optimal formation as well as network routing information. Due to the known spatial configuration of the nodes, routing is significantly simplified, since the network topology neighborhood relations are known. Routing tables can be initialized and maintained based on the knowledge of the nodes’ spatial configuration. We provide experimental results carried out with up to six nodes. Four of which are autonomous quadrotor MAVs, which are able to hover at a fixed position. One ground terminal acts as a stationary base, another ground terminal is mobile and moves. The optimal formation, given the positions of the ground stations, is automatically determined and maintained by the fleet of quadrotor MAVs. We discuss a suite of necessary further work, including automatic replacement, rearrangement and replenishment of MAVs, as well as covering user terminals at multiple locations and perching of MAVs in order to extend their battery life.