The Impact of PHY Data Link Characteristics on Cooperative Navigation for UAVs

Kurzfassung

Today, UAV swarms are a key focus of both re- search and industry. Distributed sensing, mapping, and disaster response are just a few of the use cases investigated. All of these technologies rely on precise location information for safety and coordination. While GNSS is typically used as an enabler, one needs to account for the fact that it might not always be available. Cooperative navigation based on radio messages has been proposed as a solution for this problem. This approach relies on joint communication and sensing using inter-UAV radio transmissions. The design of the respective data link determines not only the amount of data transmitted per message and packet error rate, but also the achievable measurement precision relating the states of nodes. All of these aspects influence cooperative navigation. However, while some of them have been considered in isolation, no holistic evaluation has been performed. This makes it difficult to evaluate the practical use of a cooperative navigation system based on radio transmissions. Especially when looking to simulate different algorithms for state estimation, the lack of expected data link performance values makes results less meaningful. To address this, we discuss relevant design parameters, trade-offs, and achievable performance of an OFDM data link based on the IEEE 802.11 standard. We use known signal parts to calculate Cramèr-Rao lower bounds to estimate the precision of measurements. Additionally, the number of bits that can be transmitted in a message is evaluated. Finally, we use a GNU Radio simulation to determine realistic packet error rates for different configurations. These results can be used to simulate cooperative navigation algorithms. Furthermore, we outline key trade-offs and propose data link configurations for cooperative navigation for UAVs.