Author:

Abstract:

Unmanned Aerial Vehicles (UAVs, 'drones') are being more and more widely deployed. What started as fun toys has now become a tool of the professional. Many applications quickly come to mind, for example: mapping, search and rescue and law enforcement. Current high-end UAVs are almost entirely custom designed systems. Although this type of design can deliver state-of-the-art performance, for example for the defense sector, the project can take several years. Furthermore, it leads to a high development cost that must be amortized over a usually small amount of units sold. This can keep the technology from being available to small entities with limited capital. However, only a small part of the technology in a UAV actually delivers its value. For example, all drones are able to fly, but in most cases you don't buy one for its flying abilities. Instead, you buy it for a specific application that it enables: it may come with a camera that follows you for making movies or it may have software for mapping agricultural fields. We see that these essential, but common, technologies are being more and more commoditized by community initiatives such as Ardupilot, DroneCode and Betaflight. However, communication systems are still largely proprietary and not standards based. In this thesis we show that is it possible to build wireless communication systems for UAVs from existing off-the-shelf wireless components and standardized protocol stacks. This will in many cases allow benefiting from the economies-of-scale in the portable device market. Enormous numbers of WiFi chips are sold each year, resulting in very low unit costs. However, most of these wireless technologies are not adapted for airborne use. It is shown that this can result in disappointingly low performance. Therefore, in this thesis, we are looking into ways the systems and protocol stacks can be adapted to regain the lost performance. First, the differences between traditional terrestrial communication and airborne communication are investigated. It is shown that operating at high altitudes results in propagation with less shadowing and fading effects. While this is beneficial at first sight, resulting in an extended communication range, it also leads to significantly increased interference at the airborne receiver. In addition, due to the tight integration of high-speed electronics, high-power components in the relatively small UAV platform, there is a significant risk of locally produced electromagnetic interference causing problems for airborne radios. These findings, amongst others, are shown through analysis and measurements to have a significant impact on UAV communication systems. A common UAV carries two main wireless systems. First, there is a radio that is used to connect the drone to its operator. It carries low-bandwidth command and control information. In many cases, telemetry is sent back from the UAV to the ground. In this thesis we have analyzed two off-the-shelf wireless control radios. One is a completely proprietary system, while the other solution is developed in the open source community. Both systems employ a completely custom designed air interface. It is shown, through analysis and the development of several prototype systems, that it is possible to use a standard compliant IEEE 802.15.4e system while maintaining the same level of performance. This will also simplify adding encryption and authentication to the control link, the protection of which is lacking today. The second wireless link is used for transferring payload data to the ground. While a plethora of methods are employed in commercial systems, two wireless systems are prevalent in the academic literature: WiFi and 4G LTE. Since the performance of WiFi has been shown to be less than expected in the UAV use case, we propose a system that makes it possible to deliver highly reliable video footage over IEEE 802.11. In addition, we have done a study though simulations and measurements, employing both a UAV and a small aircraft, to investigate the suitability of LTE for drone usage. Although the performance is quite good, there is one major point of interest: the enhanced propagation of the airborne UE results in significantly increased inter-cell uplink interference. This can degrade the performance of the ground network if airborne UEs are deployed, especially since UAVs, as a sensor platform, will be mostly uploading data to the network. Finally, many UAVs are critically dependent on satellite navigation systems. Loss or degradation of these signals will jeopardize the mission or even the entire drone. Therefore, it is critical to have a backup system for navigation. In this work we present a system that allows the UAV to return to its ground station, even in a GNSS degraded environment.