An enhanced multilevel ML-DFT codebook algorithm for hybrid beamforming of millimetre wave MIMO systems

Millimetre Wave (mmWave) wireless communication is considered an enabling technology to allow 5G cellular achieving high data rates. Large-scale antenna arrays can be adopted to compensate the huge path loss at higher frequencies. MIMO precoding cannot be performed only at baseband due to high power consumption of signal mixers and analogue-to-digital converters. Therefore, hybrid analogue-digital architecture at transceiver is considered as a cost-effective precoding scheme. However, the optimal design of such hybrid precoders and combiners needs to be further investigated. In this paper, a maximum likelihood (ML) beamforming technique is used for estimating signal's direction of departure in the presence of random noise. Moreover, an enhanced Orthogonal Mapping-based Matching Pursuit (OMBMP) algorithm is proposed. Layered orthogonal codebook is used to adjust base stations beamformers. Simulation results demonstrate that the proposed architecture provides acceptable performance gain with almost 90% of the performance of optimal full-digital precoder with great reduced complexity. © 2020 Inderscience Enterprises Ltd.