A Comparison of Channel Prediction Models for TDD Systems with Imperfect Channel Reciprocity

We introduce a deep learning-based channel prediction model to deal with the challenges from imperfect channel reciprocity in time-division duplex systems. We suggest several deep learning models which not only predicts future downlink channels but also calibrates channel difference between uplink and downlink channels. We also compare the performance of the introduced channel prediction models through the simulation in a wireless communication scenario.