Application of RBF neural networks to the prediction of propagation loss over irregular terrain

Point-to-point propagation models overcome the lack of accuracy of empirical models, at the expense of a large computation cost. Neural networks (NN's) provide a half-way approach to propagation modelling. In this article, it is shown that RBF NN do provide a fast point-to-point prediction method. Results show that gain in computation time is achieved while maintaining a good performance.