In the 5th generation (5G) networks, coordinated multiple point (CoMP) is one of key technologies to improve the quality of service (QoS) of edge users. To meet the requirement of growing data rates, millimeter-wave (mmWave) can be employed in the CoMP system. However, the QoS of users may be degraded if line-of-sight (LoS) mmWave channels are not guaranteed. In this article, an unmanned aerial vehicle (UAV)-aided communication scheme is proposed to enhance the QoS of edge users, where the UAV helps a primary base station (BS) and a coordinated BS simultaneously. In the proposed scheme, since the UAV only feeds back the channel state information (CSI) to the primary BS, the CSI obtained at the coordinated BS through a backbone network becomes outdated. In order to overcome the performance loss caused by the CSI feedback delay, a machine learning based channel estimation scheme is studied for the coordinated BS to perform hybrid beamforming. Furthermore, to eliminate the inter-BS interference, a maximize signal to interference-plus-noise ratio (Max-SINR) based beamforming compensation scheme is proposed for the primary BS and UAV. The simulation results show that both the bit error rate (BER) and sum rate performance can be improved by employing the proposed schemes.

中文翻译：

---

基于GPML的信道估计在无人机辅助毫米波通信中的协调波束成形

在第五代（5G）网络中，协作多点（CoMP）是提高边缘用户服务质量（QoS）的关键技术之一。为了满足不断增长的数据速率的需求，可以在CoMP系统中采用毫米波（mmWave）。但是，如果不能保证视线（LoS）mmWave信道，则用户的QoS可能会降低。在本文中，提出了一种无人飞行器（UAV）辅助的通信方案，以增强边缘用户的QoS，其中UAV同时帮助主基站（BS）和协作式BS。在提出的方案中，由于UAV仅将信道状态信息（CSI）反馈给主BS，所以通过骨干网在协作BS处获得的CSI已经过时。为了克服CSI反馈延迟导致的性能损失，研究了基于机器学习的信道估计方案，以供协作基站执行混合波束成形。此外，为了消除BS间干扰，针对主BS和UAV提出了基于最大信噪比（Max-SINR）的波束成形补偿方案。仿真结果表明，采用所提出的方案可以同时提高误码率和总和率性能。