低空密集异构网络智能干扰转移技术研究

Through fully exploiting three-dimensional spatial reuse of available spectral resources, Unmanned Aerial Vehicles (UAV)-enabled ultra-dense networks is capable of significantly increasing the capacity of wireless networks. However, due to the non-uniform deployment of UAV, the network capacity gain is limited. In view of this, this project aims to shed light on the fundamental theory of network interference uniformity and network capacity, and balance the network interference to get the interference diversity gain using the high mobility of UAV. Specifically, this project first model the uniformity of network interference considering the spatial interdependence of UAV. Then, the impact of interference uniformity on network capacity and network capacity scaling law are investigated. Furthermore, this project predicts the traffic in different region exploiting deep learning method and then design the trajectory for the UAVs to balance the network interference, get the interference diversity gain, and improve the network capacity. The achievements of this project will provide theoretical and technical supports for the application of UAVs in future mobile communications networks.

低空密集异构网络通过低空无人机搭载通信设备与地面密集化部署的基础设施进行快速、灵活组网，实现网络资源高效三维空分复用，提升网络容量。然而，低空无人机的非均匀部署导致网络干扰空间分布非均匀，从而制约网络容量增益。本项目旨在揭示低空密集网络干扰非均匀分布及其对网络容量的制约关系，并利用低空无人机的高速、有向移动性平衡网络干扰分布，获得网络干扰空间分集增益，从而使网络容量逼近其性能极限。具体而言，本项目在建模低空无人机的空间非均匀部署的前提下，刻画低空密集异构网络干扰空间分布非均匀性，研究其与网络容量的作用机理，探索网络容量随干扰非均匀程度变化趋势。在此基础上，本项目进一步采用深度学习方法预测网络业务的时空分布，并设计低空无人机的运动航线，使其与网络业务分布智能匹配，并智能平衡网络干扰分布，提升网络容量。本项目的研究为低空无人机在下一代移动通信系统中的大规模使用提供理论依据和技术支撑。

面向低空密集异构网络，本项目在准确刻画干扰空间非均匀分布的基础上，系统地研究低空密集异构网络网络容量，揭示了无人机密度和飞行高度对异构网络容量的影响，获得了最优无人机飞行高度上限和无人机临界密度；并在此基础上，设计了知识驱动的粒子群无人机轨迹优化方案，首先基于无线业务特征对网络业务进行快速精准预测，并基于预测结果及理论分析最优无人机部署位置进行快速无人机轨迹优化，从而实现与网络业务流量适配的无人机动态航线规划，为低空密集异构网络在下一代移动通信系统的应用提供关键的理论基础和技术支撑。