Capacity Analysis And Trajectory Design Methods In Unmanned Aerial Vehicle-Enabled Communication Networks

Coverage is one of the key issues that wireless networks need to solve.In order to realize the vision of the global and efficient coverage in the sixth generation mobile communication system(6G),it is urgent to solve the worldwide coverage problem,including oceans,deserts,and remote areas where are lack of network coverage,and the difficulty in providing on-demand coverage in emergency scenarios.With the help of the advantages of the unmanned aerial vehicle(UAV)-enabled communication network,including wide coverage,on-demand deployment,and flexible networking,the coverage efficiency can be improved while the network cost can be reduced.Therefore,the UAV-enabled communication network has been regarded as the significant part of the 6G wireless network.In the UAV-enabled communication network,on the one hand,compared with ground base stations(BSs),UAV nodes are sparse and heterogeneous,and the relay characteristics brought about by the aerial deployment,which makes it difficult to unify the networking model of the UAV-enabled communication network and let the backhaul connectivity be easily constrained.Therefore,the capacity performance of the UAV-enabled communication network is difficult to analyze.On the other hand,for mobile user scenarios,the UAVenabled communication network needs to follow the mobile users and continue to provide them services that meet certain performance requirements,which is referred to as the network following groups movement(NFGM).Due to the random movement of users,the lack of the prior information on the users’ location and channel state information(CSI)in future time slots,and the kinematic constraints of the UAV,all of which make the existing trajectory design methods hard to work.In addition,in the NFGM scenarios,it is difficult for the UAV-enabled communication network to obtain accurate users’ location information,and the transmission power of user terminals is limited.These factors make the uplink capacity of the UAV-enabled communication network easily constrained.In response to the above challenges,the main work and innovations of this thesis are summarized as follows:(1)To address the challenges in unifying different networking models of UAV-enabled communication networks,the easily limited backhaul links,and the difficulties in capacity analysis,this thesis proposes a networking model for the UAV-enabled communication network and analyzes the capacity of the UAV-enabled communication network based the proposed networking model.Specifically,the networking model is based on cooperative networking of the tethered UAV and the untethered UAVs,in which the tethered UAV provides aerial backhaul accesses for the untethered UAVs,while the antenna downtilt is deployed to alleviate the interference among the UAV-mounted BS cells.Based on stochastic geometry theory,the Binominal Possion Process(BPP)is utilized to model the position distribution of the untethered UAVs,and the expressions of coverage probability and ergodic capacity performance at any location in the UAV-enabled communication network based on this networking model are derived.The impacts of the backhaul connectivity,antenna downtilt,and the number of untethered UAVs on the capacity performance are analyzed.Simulation results verify the correctness of the theoretical derivation,while the obtained results highlight the advantages of the proposed networking model in terms of average ergodic rate,which can provide a reference for the corresponding network performance of other networking models.(2)To tackle the challenges in achieving the trade-off between the network capacity and coverage and dealing with the difficulties in optimizing the downlink capacity in the NFGM scenarios,this thesis proposes a UAV-enabled communication network trajectory design method oriented to the requirement of balanced network capacity and coverage.Specifically,this method adopts the directional antenna mode and defines the coverage ratio of the users’ location as a metric to evaluate the coverage performance.Meanwhile,this method transforms the trade-off problem between the network capacity and coverage into an optimization problem of maximizing the network downlink capacity with guaranteeing the coverage performance.To solve the problem,first,a global off-line three-dimensional(3-D)trajectory design algorithm is designed.Then,to solve the problems of the timeliness and the lack of prior information,a method utilizing known position information and CSI to optimize the 3-D way-point time-slot-by-slot is proposed.In addition,a comprehensive algorithm is proposed,which consists of the two-dimensional(2-D)horizontal position determination based on the density-based clustering combined with the dynamic weighted graph method,and the height optimization,as well as the timely trajectory amendment if it is necessary.The analysis results show that when the elevation angle between the scheduled user who is the farthest from the UAV is larger than a certain threshold,there exists an optimal 3-D way-point to maximize the network capacity in the corresponding time slot.The simulation results show that the proposed method can achieve the comparable average downlink throughput performance when compared with the global off-line scheme while the coverage performance is guaranteed,with approximately linear computational complexity.(3)To overcome the challenges that the network uplink capacity is easily constrained in the NFGM scenarios,where the accurate users’ location information is difficult to obtain and the uplink transmission power of user terminals is limited,this thesis proposes a radio frequency(RF)signal sensing collaborative trajectory design method for the UAV-enabled communication network.Specifically,the UAV-enabled communication network estimates users’ location information and CSI based on RF signal sensing.Then,the UAV-enabled communication network optimizes the uplink hybrid receiving beamforming and trajectory design to maximize the weighted average throughput(WAT)to improve the uplink capacity of the network.Considering that the optimization problem is NP-hard,the method first simplifies the original problem into a slot-by-slot optimization problem based on the greedy policy,and then designs algorithms based on semi-positive definite relaxation method,manifold optimization theory and successive convex approximation technology,and the alternate optimization method to obtain the suboptimal solution to the original problem.Simulation results show that in the absence of accurate location information,the proposed trajectory design method can achieve more than 95%of the WAT performance in the case of ideal location information and CSI when mobile users follow the 2-D distribution.