Transceiver Design with UCD-Based Hybrid Beamforming for Millimeter Wave Massive MIMO

Yuxing Lin,Shi Jin,Michail Matthaiou,Xiaohu You
DOI: https://doi.org/10.1109/tcomm.2019.2901690
IF: 6.166
2019-01-01
IEEE Transactions on Communications
Abstract:Hybrid transceiver designs for millimeter wave massive multiple-input multiple-output systems are feasible candidates to reduce the volume of radio frequency (RF) chains, decomposing the signal processing into the analog and digital domains. The existing schemes heavily depend on the singular value decomposition to obtain subchannels with uneven power gains, causing bit error rate (BER) performance degradation. In this paper, we propose a hybrid transceiver design based on the uniform channel decomposition (UCD), yielding subchannels with identical gains to improve the BER performance. Inspired by the UCD concept, we derive an equivalent optimization problem and propose two schemes, namely, phase-extraction and iterative update, to determine the RF beamformers, yielding an effective baseband channel with the greatest possible geometric mean of singular values. We apply the UCD with a minimum mean squared error criterion to complete the baseband beamforming. Finally, we combine the hybrid UCD beamforming with the vertical-Bell Labs layered space-time and dirty paper coding, to eliminate the inter-subchannel interference. An asymptotic analysis of the scheme performance is also pursued. The simulation results show that the proposed hybrid scheme outperforms the conventional schemes on the transmission BER, which achieves a spectral efficiency close to that of the fully-digital counterpart.
What problem does this paper attempt to address?