Efficient physics‐based recurrent neural network model for radio wave propagation in tunnels at 2.4 GHz

Abstract

Abstract The vector parabolic equation (VPE) method has been widely applied to the modelling of radio wave propagation in tunnel environments. However, dense spatial discretizations are generally required for VPE simulations to achieve acceptable accuracy. This results in high computational costs for large‐size tunnels and high‐frequency wireless systems. This letter presents an efficient physics‐based recurrent neural network (RNN) model for modeling radio wave propagation in tunnels. The proposed RNN‐based model can leverage coarse‐mesh VPE to generate high‐fidelity received signal strength and achieve significant computational savings. The validity of the proposed approach is demonstrated against fine‐mesh VPE simulations in various tunnel cases.

Keywords

• electromagnetic wave propagation
• parabolic equations
• radiowave propagation
• recurrent neural nets
• railway communication