Abstract
For high data rate wireless communication systems, developing an efficientchannel estimation approach is extremely vital for channel detection and signalrecovery. With the trend of high-mobility wireless communications betweenvehicles and vehicles-to-infrastructure (V2I), V2I communications poseadditional challenges to obtaining real-time channel measurements. Deeplearning (DL) techniques, in this context, offer learning ability andoptimization capability that can approximate many kinds of functions. In thispaper, we develop a DL-based channel prediction method to estimate channelresponses for V2I communications. We have demonstrated how fast neural networkscan learn V2I channel properties and the changing trend. The network is trainedwith a series of channel responses and known pilots, which then speculates thenext channel response based on the acquired knowledge. The predicted channel isthen used to evaluate the system performance.