Deep Learning based Power Delay Profile Trend Generation: A 60 GHz Intra-Vehicle Case Study

In this article we have utilized deep learning (DL) for channel sounding application in the millimeter wave (mmWave) band. Using data from a channel sounding campaign for studying intra-vehicle wireless channels operating over the 55-65 GHz mmWave band, we have trained an artificial neural network (ANN) model, which is used to simulate power-delay-profile (PDP) trends. The required simulation inputs form a minimal set, only comprising the frequency points, the transmitter-receiver distance and the presence of passengers inside car. The simulated PDP trend shows good match with the measured PDP and can be used for constructing tapped-delay-line (TDL) based channel models.