Artificial Neural Network for Estimating Millimeter Wave Channel Sounding Data inside a Bus

Abstract

Radio wave propagation in an intra-vehicular environment is markedly different from other well studied indoor scenarios such as an office or a factory oor. Millimeter Wave (mmWave) based intra-vehicular communications promises large bandwidth and can achieve ultra-high data rate with lower latency. However, exploiting the advantages of mmWave communications largely relies on proper characterization of the propagation channel. Channel characterization is most accurately done through an extensive channel sounding, but due to hardware and environmental constraints, it is impractical to test channel condition for all possible transmitter and receiver locations. In this paper, we use artificial neural network to aid channel sounding. Based on some real-world sounding data we show that it is possible to accurately estimate channel transfer function (CTF) and power delay profile (PDP) in an intra-bus scenario. Such artificially generated models can help in extrapolation in other relevant scenarios for which measurement data is unavailable. The proposed model can also be used for tapped delay line based bit-error-simulations as well.