Hybrid multiple-antenna transceivers, which combine large-dimensional analog pre/postprocessing with lowerdimensional digital processing, are the most promising approach for reducing the hardware cost and training overhead in massive MIMO systems. This paper provides a comprehensive survey of the various incarnations of such structures that have been proposed in the literature. We provide a taxonomy in terms of the required channel state information (CSI), namely whether the processing adapts to the instantaneous or the average (second-order) CSI; while the former provides somewhat better signal-to-noise and interference ratio, the latter has much lower overhead for CSI acquisition. We furthermore distinguish hardware structures of different complexities. Finally, we point out the special design aspects for operation at millimeter-wave frequencies. Index TermsHybrid beamforming, 5G, Millimeter-wave.

This work studies a variant of hybrid beamforming, namely, hybrid beamforming with selection (HBwS), as an attractive solution to reduce the hardware cost of multi-user Massive Multiple-Input-Multiple-Output systems, while retaining good performance. Unlike conventional hybrid beamforming, in a transceiver with HBwS, the antenna array is fed by an analog beamforming matrix withL input ports, whereL is larger than the number of up/down-conversion chainsK. A bank of switches connects the instantaneously bestK out of theL input ports to the up/down-conversion chains. The analog beamformer is designed based on average channel statistics and therefore needs to be updated only infrequently, while the switches operate based on instantaneous channel knowledge. HBwS allows use of simpler hardware in the beamformer that only need to adjust to the statistics, while also enabling the effective analog beams to adapt to the instantaneous channel variations via switching. This provides better user separability, beamforming gain, and/or simpler hardware than some conventional hybrid schemes. In this work, a novel design for the analog beamformer is derived and approaches to reduce the hardware and computational cost of a multi-user HBwS system are explored. In addition, we study howL, the switch bank architecture, the number of users and the channel estimation overhead impact system performance.

No abstract

No abstract