Improved Maritime Channel Model and Channel Predictor Utilizing Transceiver Pitch Motions

In this letter, a sea wave-oriented geometry-based statistic model (SW-GBSM) is proposed for the maritime communication scenario. Different from the existing model that only considers the heave motion of transceiver antennas, the proposed channel model further takes into account their pitch motions due to the fluctuating sea waves. As a result, the small-scale fading and Doppler shift effects of the maritime channel can be better characterized. Moreover, a sea wave feature-based gated recurrent unit (SWF-GRU) network is introduced for prediction of the sea surface channel. The specific feature extracted from the pitch motions of the transceiver enables the network to adapt to the varying Doppler shifts caused by the fluctuating sea waves, thus contributing to an improved channel prediction accuracy. Simulation results show that incorporating the pitch motion aspect makes the channel model more precise and the channel predictor more efficient.