Abstract
The impact of incidence angle on radio wave propagation prediction has been studied and properly modeled empirical approach and equation for a long time. Obtained results are mitigated about the consideration of these parameters in radio wave prediction since the gain is few dB while introducing more computing complexity. In this paper we are interesting in studying the impact of the incidence angle on final predicted results when using artificial neural network (ANN) approach. Our work consists at the beginning in developing a neural model which inspires its inputs from Cheung model and integrate the incidence angle between wave and obstacle as input parameter. Models accuracy will be estimated basing on the comparison between real collected measurements and model predictions. Obtained results are than compared to our previous published work which consist in developing a neural model inspired from Multi-Wall which not incorporate the incidence angle. This comparison has showed that the consideration of incidence angle improves slightly the prediction accuracy that can be neglected due the complexity and computational time that it introduces.