Massive Multiple-Input Multiple-Output (MIMO) technology has the potential to deliver high spectral efficiencies by leveraging the ability to form narrow beams that can be resolve users based on location. Complexity is a significant barrier, particularly for massive MIMO systems employing orthogonal frequency division multiplexing (OFDM). This paper presents a simple technique for employing multiple-beamforming on the downlink of massive MIMO-OFDM systems. With a focus on practicability, the channel state information (CSI) is specified with only two scalar parameters, namely, beam angle and average channel signal to noise ratio. A novel channel estimation procedure is described, wherein, all mobile users will acquire CSI simultaneously using circularly-shifted pseudo-noise training sequences. The base station then performs an optimization task to identify suitable spectrum (or subcarriers) for each user. We have shown an iterative algorithm for orthogonal multiple access in this paper. Extension to non-orthogonal multiple access is straightforward. Simulation results are presented to show the behavior of our multiple-beam OFDM technique. We have considered only the cellular downlink in this paper.

大规模多输入多输出（MIMO）技术具有利用形成窄波束的能力来提供高频谱效率的潜力，这些波束可以根据位置解析用户。复杂性是一个重大障碍，特别是对于采用正交频分复用（OFDM）的大规模MIMO系统而言。本文提出了一种在大规模MIMO-OFDM系统的下行链路上采用多波束成形的简单技术。着眼于实用性，仅用两个标量参数（即波束角和平均信道信噪比）来指定信道状态信息（CSI）。描述了一种新颖的信道估计过程，其中，所有移动用户将使用循环移位伪噪声训练序列同时获取CSI。然后，基站执行优化任务，以为每个用户标识合适的频谱（或子载波）。在本文中，我们展示了一种用于正交多址访问的迭代算法。扩展到非正交多址访问非常简单。仿真结果表明了我们的多波束OFDM技术的性能。在本文中，我们仅考虑了蜂窝下行链路。