Optimization Of Spectrum-and Energy-Efficiency Of UAV-BS Assisted Communication Systems

The Fifth Generation Mobile Communications(5G)place increasing demand for high data rate thus it is necessary to enhance spectrum efficiency(SE).Unmanned Aerial Vehicle(UAV)has become a new research hotspot in 5G and attracted much attention from researchers.In contrast to terrestrial infrastructure in cellular networks,UAV has the advantages of fast deployment and low cost as well as strong Line of Sight(Lo S)component in air-to-ground link.However,the main drawback of UAV is limited battery endurance.Motivated by these issues,this dissertation studies SE and EE optimizations for UAV base station(BS)aided communication with two-dimensional(2D)and three-dimensional(3D)channel model,respectively.With 2D trajectory model,this dissertation considers a multi-antenna UAV BS implementing directional beam to cover ground users via angle-domain 3D beamforming.Based on practical power consumption model involve d with propulsion and communication power,EE and SE are studied.It can be found that with a given transmission power which is normally less than the propulsion power,high EE can be maintained by maximizing SE.Hence,the joint SE optimization problem is formulated.To tackle it,a new algorithm is proposed to divide the original problem into four subproblems as user scheduling,power allocation,beamforming tilt optimization and UAV 2D trajectory design.Simulation results show that 3D BF plays a dominant role in the overall performance,and the proposed closed-form 3D BF design achieves near-optimal performance in terms of both EE and SE,while requiring much lower complexity than exhaustive search.This dissertation further considers ground users are served by a single antenna UAV BS and optimizes EE as well as SE with 3D UAV trajectory model.Assuming flexible flying altitude,the air-to-ground link is modeled as time-varying Rician channel.A joint optimization problem for UAV 3D trajectory and user scheduling is formulated.Simulation results show that the 3D trajectory design achieves much performance enhancement over 2D trajectory design in terms of EE and SE.