A Res-FCNN-Based Correction Method of Urban Environment Radio Propagation Model

Radio wave propagation models play a crucial role in the design and optimization of communication systems, especially in complex urban environments. However, traditional radio wave propagation models often exhibit deviations in practical applications, necessitating correction to improve prediction accuracy. Firstly, this paper proposes a method for correcting radio wave propagation models in urban environments and verifies its feasibility using the ITU-R P.1411 model as an example. Secondly, an urban environment radio wave propagation model was constructed based on the ITU-R P.1411-12 recommendation. Finally, the Residual Fully Connected Neural Network (Res-FCNN) model was used to correct it. Simulation results show that the prediction accuracy of the ITU-R P.1411 model is significantly improved after deep learning correction.