海上无人机基站高能效毫米波传输技术研究

Due to the benefits of good mobility, quick deployment, low cost and easy to upgrade, the unmanned aerial vehicle (UAV) plays an important role in constructing the ocean observing networks so that our country can achieve the change from a big oceanic country to a strong one. The goal of this project is to improve the capacity of data transmission in the ocean observing networks by focusing on the energy efficient millimeter-wave (mmw) transmission for UAV. Firstly, we will design a channel model of mmw based on the characteristic of UAV and ocean environment and analyze the energy efficiency and corresponding coverage area for the mmw transmission, revealing the effect of key factors on the energy efficiency. Then, the optimal and sub-optimal beamforming schemes will be designed to maximize energy efficiency and an adaptive selection scheme based on reinforcement learning will also be proposed to trade-off the system performance and the computational complexity. Finally, we will focus on solve the energy efficiency maximization problem with imperfect channel state information (CSI). By training from the imperfect CSI samples, a robust beamforming scheme based on the semi-supervised learning will be presented to improve the robustness of mmw transmission. The innovative works of this project will provide the theoretical and technical approaches to solve some bottleneck problems which restrict the development of UAV mmw transmission technologies, and promote its practical application in the ocean observing networks.

无人机基站作为通信平台具有机动性好、部署快、成本低以及易于系统升级等优势，对于我国构建海洋观测网，实现由海洋大国向海洋强国的转变具有重要的意义。本项目以提升海洋观测网信息传输容量为目标，重点研究基于能量效率的无人机基站毫米波传输方法。通过建立综合体现无人机基站海上通信特点的毫米波信道模型，提出系统能量效率及其覆盖区域的分析方法，揭示影响系统能量效率的主导因素和其作用机理；基于此，探索能量效率最大化准则下的最优和次优波束成形算法及基于强化学习的自适应传输选择策略，提升传输能效的同时实现性能与复杂度的最佳折中；最后，着力解决信道状态信息存在误差下的能效优化问题，通过对信道误差的训练学习，提出基于半监督回归学习的鲁棒波束成形算法，增强毫米波传输对误差的容忍能力。本项目的创新研究将会为无人机基站毫米波传输技术的发展提供理论依据和技术方案，推动其在我国海洋观测网中的实用化进程。

随着无人机应用需求带来的无线数据传输容量爆发式增长，无人机通信系统仍面临着机载能量有限和无线信道窃听两大难题。本项目围绕无人机通信系统中亟待解决的能效性和安全性问题开展了深入的研究，主要研究内容及学术贡献如下：（1）针对无人机空地信道直达链路和散射链路随飞行高度的变化特性，建立了无人机空-地信道概率模型，提出了无人机最佳高度计算方法和部署方案，在满足地面用户通信覆盖要求下可有效降低无人机通信功耗，提升能效性。（2）针对多用户通信的安全性需求，给出了存在非法窃听场景下的无人机安全能效定义，提出了基于多用户安全能效公平性的无人机通信系统资源分配与飞行航迹联合优化算法，可有效保障无人机作业期内多用户安全能效性能的公平性。（3）针对无人机抖动特性，建立了无人机抖动误差模型，提出了基于安全能效指标的无人机毫米波通信系统鲁棒波束成形优化算法，可有效保障无人机抖动情况下的数据传输性能。（4）针对全双工毫米波无人机中继系统，建立了基于毫米波平面阵的无人机空-地信道及无人机全双工自干扰信道模型，并提出了以无人机通信安全能效为优化指标的无人机位置和波束成形联合优化算法，实现无人机最佳中继部署与安全能效优化。.在三年的研究时间内，本项目共投入研究人员17人。经过课题组的齐心协力和刻苦攻关，共发表学术论文27篇，其中IEEE期刊论文11篇；申请国家发明专利10件，授权2件；培养研究生13名。.本项目的创新性研究为尽快解决制约无人机通信系统发展的若干瓶颈问题提供理论依据和技术方案，推动其在民用和军事无线通信领域的应用。