Beam Selection for Beam Refinement Procedure in Multi-cell Millimeter-Wave Massive MIMO Systems

Abstract

The millimeter wave (mmWave) communications have gained increasing attention to satisfy the high data-rate requirements in the fifth generation (5G) cellular systems. To overcome the severe path loss, multiple narrow beamforming vectors are configured at both the next generation NodeB (gNB) and the user equipment (UE), thus it is critical to select the best beam pair for maximum beamforming gain. This paper focuses on the beam selection problem for beam refinement procedure, where selection occurs between each UE with its serving gNB. A hybrid approach is proposed where the beam pair is selected either with the max-signal-to-noise (Max-SNR) criteria or with the max-signal-to-interference-plus-noise ratio (Max-SINR) criteria, depends on the highest reference signal received power (RSRP) measured by UE. Simulation results indicate that the proposed hybrid approach can achieve a good tradeoff between the beam selection delay and the SINR performance