本文为芬兰奥卢大学（作者：SivalingamThushan）的硕士论文，共43页。

当通信主干网在临时事件和灾难发生后无法提供覆盖和容量需求时，无人机基站是一种可靠而有效的替代方案。在这篇论文中，我们考虑多个无人机基站在毫米波网路中的三维部署。首先，我们定义了一组UAV-BS部署在一个小区内的位置，然后确定了多个UAV BSs的预定义位置的可能组合，并假设用户具有五个固定的位置。我们提出了一种新的算法来从多个无人机预定位置中找出可行的位置，以保证服务质量（QoS），由于管制政策，无人机BS的悬停高度和作业区域都会受到限制。此外，我们考虑毫米波传输和多天线技术来产生定向波束以服务于小区内的用户。我们把定位问题归结为一个l0最小化问题。这是一个组合、NP-hard问题，最优解不可通过穷举搜索来处理。因此，我们着重研究次优算法来寻找可行解。我们将极小化问题近似为非组合l1范数问题。仿真结果表明，利用毫米波传输，满足约束条件的无人机基站定位是可行的。进一步的分析表明，该算法可以达到一个接近最优的位置，以实现同时部署多个UVA基站。

Unmanned aerial vehicle (UAV) base stations(BSs) can be a reliable and effiffifficient alternative to full fifill thecoverage and capacity requirements when the backbone network fails to providethe requirements during temporary events and after disasters. In this thesis,we consider three-dimensional deployment of multiple UAV-BSs in amillimeter-Wave network. Initially, we defifined a set of locations for aUAV-BS to be deployed inside a cell, then possible combinations of predefifinedlocations for multiple UAV-BSs are determined and assumed that users havefifixed locations. We developed a novel algorithm to fifind the feasiblepositions from the predefifined locations of multiple UAVs subject to asignal-to-interference-plus-noise ratio (SINR) constraint of every associateduser to guarantees the quality-of-service (QoS), UAV-BS’s limited hoveringaltitude constraint and restricted operating zone because of regulationpolicies. Further, we take into consideration the millimeter-wave transmissionand multi-antenna techniques to generate directional beams to serve the usersin a cell. We cast the positioning problem as an `0 minimization problem. This is a combinatorial, NP-hard, andfifinding the optimum solution is not tractable by exhaustive search. Therefore,we focused on the sub-optimal algorithm to fifind a feasible solution. Weapproximate the minimization problem as non-combinatorial `1-norm problem. The simulation results reveal that, withmillimeter-wave transmission the positioning of the UAV-BS while satisfying theconstrains is feasible. Further, the analysis shows that the proposed algorithmachieves a near-optimal location to deploy multiple UVA BS simultaneously.

1. 引言

2. 项目背景

3. 多无人机基站定位

4. 仿真结果与讨论

5. 结论与展望

更多精彩文章请关注公众号：