The emerging millimeter-wave (mmWave) massive multiple-input multiple-output
(MIMO) with lens antenna arrays, which is also known as "beamspace MIMO", can
effectively reduce the required number of power-hungry radio frequency (RF)
chains. Therefore, it has been considered as a promising technique for the
upcoming 5G communications and beyond. However, most current studies on
beamspace MIMO have not taken into account the important power leakage problem
in beamspace channels, which possibly leads to a significant degradation in the
signal-to-noise ratio (SNR) and the system sum-rate. To this end, we propose a
beam aligning precoding method to handle the power leakage problem in this
paper. Firstly, a phase shifter network (PSN) structure is proposed, which
enables each RF chain in beamspace MIMO to select multiple beams to collect the
leakage power. Then, a rotation-based precoding algorithm is designed based on
the proposed PSN structure, which aligns the channel gains of the selected
beams towards the same direction for maximizing the received SNR at each user.
Furthermore, we reveal some system design insights by analyzing the sum-rate
and energy efficiency (EE) of the proposed beam aligning precoding method. In
simulations, the proposed approach is found to achieve the near-optimal
sum-rate performance compared with the ideal case of no power leakage, and
obtains a higher EE than the existing schemes with either a linear or planar
array.

本文针对毫米波巨型多输入多输出 (beamspace MIMO) 技术中的功率泄露问题，提出了基于相移器网络 (PSN) 和旋转预编码算法的波束精准对齐方法，使所选波束具有相同的接收信噪比以最大化用户的接收信噪比。我们的研究显示，该方法在近似无功率泄露情况下，sum-rate 性能几乎达到最优，并且比现有的具有线性或平面阵列的方案获得更高的能量效率 (EE)。