Service Outage Estimation for Unmanned Aerial Vehicles: A Measurement-Based Approach

One of the key enablers for the use of cellular networks to provide connectivity to Unmanned Aerial Vehicles (UAVs) is reliability assurance. Estimating accurately the expected service availability and reliability that a UAV will experience along a planned route, helps to avoid critical situations. In this paper, we evaluate the use of a data-driven approach to estimate the expected serving signal level of a UAV planning to fly along a specific path. The estimation approach, presented in a previous study for the ground-level case, is based on the aggregation of measurements of UAVs that have previously been flying along the same route. Using this approach we achieve an estimation error as low as 2.7 dB. Based on the estimations we calculate the expected outage probability in terms of service availability and reliability and provide estimations of the expected critical areas along the route. Results show that 90% of the availability and 65% of the reliability critical areas in the route can be accurately estimated.