This article develops a novel hierarchical 5G Internet of Things network with unmanned aerial vehicles (UAVs) in the sky. In the proposed system, the leader UAV plays a vital role in the communication with ground base stations and other UAVs. The leader UAV relies on multiple beamforming to establish and maintain reliable broadband connections, which requires the location and altitude information of the UAV. Therefore, we propose a novel deep learning algorithm based on gated recurrent units and autoencoder for trajectory prediction and pose estimation. Simulation results show that this algorithm greatly improves the performance of the entire system, and has obvious advantages compared to traditional methods.

中文翻译：

---

基于深度学习的5G UAV IoT网络的多波束成形

本文使用空中无人飞行器（UAV）开发了一种新颖的5G分层物联网网络。在提出的系统中，领导者无人机在与地面基站和其他无人机的通信中起着至关重要的作用。领先的无人机依靠多种波束成形来建立和维护可靠的宽带连接，这需要无人机的位置和高度信息。因此，我们提出了一种基于门控递归单元和自动编码器的新型深度学习算法，用于轨迹预测和姿态估计。仿真结果表明，该算法大大提高了整个系统的性能，与传统方法相比具有明显的优势。