Telecommunication network coverage using autonomous drones by Thomas Wilmet

There is a growing interest in the use of UAVs (Unmanned Aerial Vehicles) across very different fields, and many applications are currently under development. Within the BEAMS department of the Université libre de Bruxelles (ULB), under the direction of François Quitin, one such project aims at using drones for network coverage.

The idea of the project is the following: developing an autonomous drone that, when released, locates the appropriate area where the drone can be used as flying relocatable cell tower, to replace or to support the existing cellular network.

A simple example of the importance of such a system is the case of a natural disaster, that would wipe out the entire 3G and 4G network. Such an event would block all wireless communications for victims and complicate rescue operations. The drone could be released autonomously, determine the area where they would be the most useful and temporarily replace the damaged wireless network.

The overall project can actually be divided in three tasks : first, the drone should locate the ground users. Let us imagine there are hundreds of ground users, it is quite intuitive that the drone must locate and fly towards the area where the population is most dense to get the best signal-to-noise ratio, resulting in the best overall performance. The second problem is determining the optimal location (or trajectory) of the drone in the area where the population is most dense. How to do this based solely on the radio signals captured by the drone is quite challenging optimization problem. The third and final problem is how to implement a reliable backhaul link between the drone and the backhaul network. Currently, it is mainly the first of these three problems that is currently being studied: how to locate a ground user based on the radio signals that the drone receives.

This is done by using four directional antennas mounted on the drone itself arranged perpendicularly, in the four cardinal directions. While flying, each antenna of the drone will continuously wait for a signal at the frequency of interest until it detects one.

The antennas oriented towards the transmitter will receive the strongest signal, and by weighing the power received at each antenna, the direction of the ground transmitter can be determined. The drone then starts moving in the estimated direction. This procedure is repeated at each iteration, until the drone gets close to transmitter.

Unfortunately drones still are a relatively recent invention. As such, autonomous drones can take time to be authorized by the legislation. For obvious matters, autonomous drone should not be put in service before the legislation finishes to establish the rules and barriers to ensure the safety of both users and civilians. To be realistic, this fact delays for about 5 to 10 years the time we will see network coverage through autonomous drones, if the project still is useful at that time.

Even though the legislation establishment is slowing down the overall project, some studies within the project could actually be re-used in a very near future. A first example would be relief operations in rough environments, where network coverage is not ensured everywhere. The drone could locate the users from their signal, which could not be done via the classical network for lack of coverage, and then could fly towards the emitter without having to inquire about how to get there via a ground track, which could be difficult in mountains for example. Another example would be the illegal use of some restricted precise frequencies of network operator, such as the IBPT e.g. Indeed, thanks to the variable frequency detection of the drone and its hooked antennas, one could set-up the antennas to look for one of the restricted frequency. If one of these is used illegally, the drone will detect it and fly towards the cyber-criminals which would help greatly the network operators that pay for these frequencies exclusivity.

So even though we have to wait to see the main project emerge, it still finds very concrete and direct applications for a near future that, if successful, could open new perspectives in a global network coverage.