Artificial Neural Network Based Simulation of Short-Term Fading in Mobile Propagation Channel

A novel method for correlated short-term fading simulation is presented and analyzed in this paper. Proposed solution is based on artificial neural networks. In order to obtain an adequate training data set, extensive measurements of the electric field strength were carried out in an indoor environment. Due to its modular architecture, the proposed simulator could be used to generate potentially any number of correlated fading processes (or a single fading process). For verification purposes, the proposed method was compared against the existing simulation methods (of both single and two correlated fading processes). The quantitative and qualitative analysis showed that the proposed method was in better agreement with the measurements comparing to the existing methods.