弱视距下用户为中心无人机群组网非典型空时性能理论分析

面对密集城区弱视距通信环境，用户为中心无人机群组网克服单无人机基站受限，颠覆基站为中心架构，实现无边界小区设计，保障业务一致性与连续性。然而群组设计引入相关性，弱视距传输加剧相关性变化，使得传统的典型用户分析方法不再适用。本课题针对现有自由空间无人机模型不符合实际的缺点，构建带有边界约束和遮挡效应的网络模型；针对典型用户分析方法忽略了相关性的缺点，将网络状态相关性的大小和特征量化表征为空时相关性因子；针对空间相关性导致的单用户性能不具有典型性问题，分析非典型空间性能随空间相关性因子的定量变化规律，体现区域统计和局部差异性；针对时间相关性导致的瞬时性能不具备遍历性问题，分析非典型时间性能与时间相关性因子的定量关联关系，表征时间统计和多时隙依赖性。基于相关性分析最终得到“简洁而紧致”的非典型空时性能理论界，从理论上证明无人机群组网增益来源，形成较为完整的B5G无人机群组网络理论基础及技术体系。

Facing the weak line-of-sight communication environment in the dense urban, the user-centric UAV cluster-assisted network shifts the traditional cell-centric design and gives the design of borderless cells to overcome the limitation of single UAV base station. Therefore, the consistency and continuity of the service are realized. However, the cluster design introduces the correlation, and the weak line-of-sight communication environment intensifies the change of the correlation, both of which makes the traditional analysis method based on typical users no longer applicable. For the defect that the existing free space UAV models are impractical, this project aims to construct UAV network model with boundary effect and blocking effect. For the defect that typical-user-based analysis method ignores the correlation, this project quantifies the degree and characteristics of the correlation of the network states as spatio-temporal correlation factors. For the issue that the single-user performance is not typical due to the spatial correlation, this project analyzes the quantitative variation of the atypical spatial performance with spatial correlation factors, reflecting spatial statistical properties and local difference characteristic. For the issue that the instantaneous performance no longer has ergodicity due to the temporal correlation, this project analyzes the quantitative variation of the atypical temporal performance with temporal correlation factors, characterizing temporal statistical properties and dependency among multiple time-slots. The compact theoretical bounds for atypical spatio-temporal performance are obtained through the correlation analysis, so as to explain theoretically the source gains of user-centric UAV cluster-assisted network. This project will build a more complete theoretical and technical system for B5G UAV cluster-assisted network.