Unmanned aerial systems (UAS) have recently gained popularity, and they are envisioned as an integral parts of the current and future wireless and mobile-radio networks. Despite the exhaustive research on air-to-ground channels, there are insufficient studies, experimental campaigns and general channel models related to air-to-space (A2S) and air-to-air (A2A) wireless links. This paper presents a comprehensive review of the available channel models and path-loss prediction for A2S and A2A communications. Specific case studies attempting to extend current models’ parameters and provide important knowledge of the channel behavior in combination with UAV flight characteristics are also provided. A time-series rain-attenuation synthesizer is also presented that describes quite accurately the impact of the troposphere at frequencies above 10 GHz. This specific model can be also applied to both A2S and A2A wireless links. Finally, scientific challenges and gaps that can be used for future research on the upcoming 6G networks are highlighted.

This letter presents the measurements of dual-polarized multiple-input-multiple-output (DP-MIMO) channel between a flying Unmanned Aerial Vehicle (UAV) and a ground terminal. The campaign is carried out in an urban area under with trees surrounding the receiver. The UAV-to-ground channel is statistically modeled by a novel proposed composite distribution that is realized from the combination of the Inverse Gaussian (IG) and Rayleigh processes, in view of characterizing the shadowing and multipath effects of the channel. It is shown that the suggested new mixed model exhibits a noticeable adaptability, fitting very well to the measured data and delivering errors below 5%, according to the Kullback–Leibler distance metric.

This paper presents an extensive performance analysis of a narrowband dual-polarized (DP) multiple-input-multiple-output (MIMO) communication link, between a low-altitude unmanned aerial vehicle (UAV) and a terrestrial terminal. The experimental campaign is conducted in an urban park, under tree shadowing environment. The ergodic capacities are calculated 5.2 b/s/Hz for line-of-sight (LOS), and 4.4 b/s/Hz for non-line-of-sight (NLOS) trajectories, respectively, substantially enhanced compared with the single-input-single-output (SISO) cases. The encountered capacity variations are associated with the propagation geometry and the UAV flight path, as well as with the elevation angle and distance between the related terminals. From the multiplexing assessment analysis the DP-MIMO channel has been shown its favourability for adopting beamforming methods. A new stochastic model is also suggested that represents the channel matrix employing jointly the Inverse Gaussian and Rayleigh distributions. Promising results are delivered, where the adopted model predicts conveniently the channel capacity under the existence of tree shadowing. Finally, shorter stationarity periods are observed under the tree presence affecting their pattern and, in respect, the achieved instantaneous capacity.