A Novel Non-Stationary 6G UAV Channel Model for Maritime Communications

TLDR

UAV communications serving as mobile base stations in maritime scenarios are increasingly important for future 6G space‑air‑ground‑sea integrated networks, and accurate UAV‑to‑ship channel modeling is essential for system design, testing, and performance evaluation. This study proposes a novel non‑stationary, multi‑mobility UAV‑to‑ship channel model that incorporates line‑of‑sight, single‑bounce, and multi‑bounce components arising from sea‑water fluctuations and waveguide effects. The model treats the UAV transmitter, ship receiver, and intermediate clusters as moving with arbitrary velocities and directions, and derives key statistical properties—including temporal autocorrelation, spatial cross‑correlation, Doppler, delay, and angular power spectra, stationary intervals, and RMS delay spread. Comparison with available measurement data confirms the accuracy of the proposed channel model.

Abstract

To achieve space-air-ground-sea integrated communication networks for future sixth generation (6G) communications, unmanned aerial vehicle (UAV) communications applying to maritime scenarios serving as mobile base stations have recently attracted more attentions. The UAV-to-ship channel modeling is the fundamental for the system design, testing, and performance evaluation of UAV communication systems in maritime scenarios. In this paper, a novel non-stationary multi-mobility UAV-to-ship channel model is proposed, consisting of three kinds of components, i.e., the line-of-sight (LoS) component, the single-bounce (SB) components resulting from the fluctuation of sea water, and multi-bounce (MB) components introduced by the waveguide effect over the sea surface. In the proposed model, the UAV as the transmitter (Tx), the ship as the receiver (Rx), and the clusters between the Tx and Rx, can be seen as moving with arbitrary velocities and arbitrary directions. Then, some typical statistical properties of the proposed UAV-to-ship channel model, including the temporal autocorrelation function (ACF), spatial cross-correlation function (CCF), Doppler power spectrum density (PSD), delay PSD, angular PSD, stationary interval, and root mean square (RMS) delay spread, are derived and investigated. Finally, by comparing with the available measurement data, the accuracy of proposed channel model is validated.