Channel and Carrier Frequency Offset Equalization for OFDM Based UAV Communications Using Deep Learning

摘要

Use of preambles and/or pilot sequences for equalizing the channel and carrier frequency offset (CFO) effects results in inefficient use of allocated bandwidth. In this work, we propose a deep learning based channel and CFO equalization technique for orthogonal frequency division multiplexed (OFDM) systems. The proposed method is data driven and doesn't make use of any preamble or pilot sequences, making it bandwidth efficient compared to the existing methods. To demonstrate the effectiveness of the proposed method, we consider OFDM based unmanned aerial vehicle (UAV) communication systems, which are prone to time varying channel and CFO effects due to their constant motion. Simulation results show that the proposed method performs better than the existing methods and works for various propagation environments.