Hybrid Beamforming for Large-scale MIMO-OFDM in Frequency Selective Fading

Hybrid analog-digital precoding is a class of efficient techniques for improving the performance of massive multiple-input multiple-output (MIMO) systems with reduced number of radio frequency (RF) chains. Meanwhile, for combating the frequency selective fading, orthogonal frequency-division multiplexing (OFDM) should be also combined in current wireless standards such as 5G. Therefore, for massive MIMO-OFDM, it is difficult to decide the best precoding matrix in the context of frequency selective fading. In this paper, for alleviating this problem, a novel hybrid analog-digital precoding algorithm is proposed for multi-user large-scale MIMO-OFDM systems when the frequency selective fading channel is considered. We show that the proposed algorithm offers a flexible precoding design as well as reducing the requirement of the number of RF chains. Finally, the performance results demonstrate that the proposed algorithm is capable of approaching the performance of optimal full-digital precoding with a low implemention complexity.