In this paper, adaptive hybrid beamforming methods are proposed for millimeter-wave range massive multiple-input-multiple-output (MIMO) systems considering single carrier wideband transmission in uplink data mode. A statistical analog beamformer is adaptively constructed in slow-time, while the channel is time-varying and erroneously estimated. A recursive filtering approach is proposed, which aims robustness against estimation errors for generalized eigen-beamformer (GEB). Approximated expressions are obtained for channel covariance matrices that decouple angular spread and center angle of multipath components. With these expressions, modified adaptive construction methods for GEB are proposed, which use only the quantized estimated power levels on angular patches. The performances of the proposed slow-time adaptation techniques for statistical Massive MIMO beamforming are evaluated in terms of the output signal-to-interference-and-noise-ratio (SINR), instantaneous channel estimation and beam accuracy. They are shown to be very efficient such that the computational complexity is significantly reduced while the performance remains almost the same as that of the ideal GEB even in large angular estimation errors.

中文翻译：

---

毫米波时变信道中大规模 MIMO 混合波束成形的高效慢时自适应

在本文中，针对上行数据模式下考虑单载波宽带传输的毫米波范围大规模多输入多输出（MIMO）系统提出了自适应混合波束成形方法。统计模拟波束形成器是在慢时间自适应构建的，而信道是随时间变化的和错误估计的。提出了一种递归滤波方法，其目的是针对广义特征波束形成器（GEB）的估计误差具有鲁棒性。获得了信道协方差矩阵的近似表达式，该矩阵解耦了多径分量的角度扩展和中心角。使用这些表达式，提出了改进的 GEB 自适应构造方法，该方法仅使用角块上的量化估计功率电平。所提出的用于统计大规模 MIMO 波束成形的慢时自适应技术的性能在输出信干噪比 (SINR)、瞬时信道估计和波束精度方面进行了评估。它们被证明是非常有效的，因此计算复杂度显着降低，而即使在较大的角度估计误差下，其性能也几乎与理想的 GEB 相同。