| In recent years,the requirement for mobile data services has grown exponential-ly.However,the current 4th-generation mobile communication systems are difficult to meet the demand for spectrum efficiency and energy efficiency of future mobile commu-nications.In order to increase wireless communication transmission rate by thousands times of existing systems in 2020,further exploration of spectrum resources and spatial multiplexing capability of multiple antennas is a key way to improve spectrum efficien-cy and energy efficiency.The millimeter-wave band has a large number of blank free resources and has not been fully developed.Therefore,millimeter-wave communication-s have been highly valued by telecom departments and enterprises in various countries.At the same time,large-scale multi-antenna systems can further exploit the spatial mul-tiplexing capability of multiple antennas,thus greatly improving spectral efficiency and energy efficiency.Therefore,the millimeter wave massive multiple-input multiple-output technology is the key technology of the future mobile communication systems,and can provide a strong guarantee for the future mobile communication systems.Considering the issues including channel physical parameter estimation,transmis-sion mechanism,multi-user scheduling,dynamic resource allocation and other issues in millimeter wave massive multiple-input multiple-output systems,and the target of im-proving spectrum efficiency,this dissertation studies channel estimation,transmission mechanism,and multi-user scheduling in different environments,different antenna struc-tures,different precoding architectures and UAV swarm scenarios by using model estab-lishment,parameter extraction,algorithm design,performance analysis and simulation verification.The main contributions of the dissertation include:1)The dissertation exploits the antenna array theory for channel estimation in 60GHz indoor massive uniform planar array communications environment.The channel es-timation can be decomposed into angular estimation and gain estimation,which is a unique property for massive multiple-input multiple-output systems.Then this dis-sertation proposes an array signal processing aided fast direction of arrival estimation algorithm,and the gain information could be obtained with very small amount of training resources.Moreover,the angle reciprocity is utilized to facilitate the down-link channel estimation for both time division duplexing and frequency division du-plexing systems.To enhance the spectral and energy efficiency,an angle division multiple access user scheduling algorithm based on angular information of different users is designed.Compared to existing channel estimation algorithms,the newly proposed one does not require any information of channel statistics,can be efficiently deployed by the two dimensional discrete Fourier transform operations,and is unified for both time division duplexing and frequency division duplexing systems,making it a practical solutions for 60GHz indoor communications.2)The dissertation proposes a novel channel estimation for hybrid digital and analog millimeter wave massive multiple-input multiple-output system,where the channel is decomposed into direction of arrival information and channel gain information.In the proposed estimation method,a fast direction of arrival estimation algorithm was designed based on two dimensional discrete Fourier transform operation and angle rotation technology,and the channel gain estimation can be performed with very small amount of training resources.To evaluate the benefits of our proposed method,the theoretical bounds of mean square error and Cramer-Rao lower bound of the joint direction of arrival and channel gain estimation in high signal-to-noise ratio region are derived.It is shown that our proposed estimation method is very close to the Cramer-Rao lower bound,especially in the large base station antennas case.3)The dissertation investigates the problems of channel estimation,hybrid precoding,user scheduling,and power allocation for millimeter wave non-orthogonal multiple access system with hybrid architecture.By utilizing the special structural character-istics of millimeter wave channel,this dissertation proposes a general iterative index detection-based channel estimation algorithm both estimating direction of arrival and channel gain for each channel path.Then this dissertation proposes an angle domain hybrid precoding scheme to reduce the inter-beam interferences.With the objective of maximizing the system achievable sum rate,a non-convex problem that jointly optimizes user scheduling and power allocation is formulated with the interference constraints of different users.To solve this non-convex problem,this dissertation decomposes it into two sub-problems,i.e.,user scheduling and power allocation sub-problems.A novel matching theory based user scheduling algorithm is designed to solve the user scheduling sub-problem.With the user scheduling results,an iter-ative optimization algorithm is developed to realize the dynamic power allocation.Simulation results show that the proposed direction of arrival estimation and channel estimation outperform a better mean square error performance compared with con-ventional methods.Moreover,the sum rate of the proposed matching theory based user scheduling algorithm and power allocation schemes also outperform the conven-tional millimeter wave beamspace multiple-input multiple-output system.4)The dissertation proposes a new channel estimation and self-positioning method for the unmanned aerial vehicle swarm where the channel is decomposed into direction of arrival information,the relative position information and the corresponding chan-nel gain information.This dissertation proves the uniqueness of the null space of the determinant matrix.The channel gain estimation is performed with very small amount of training resources.Theoretically,the ambiguity of positioning estimation does not exist when the number of the transmitted signals is greater than or equal to 3.However,outlier may happen in the practical estimation due to the existence of the noise.This dissertation develops two efficient algorithms for the positioning esti-mation that avoid the exhaustive searching over the entire lattice region.Meanwhile,several ways that could improve the detection accuracy are suggested.Moreover,the deterministic Cramer-Rao lower bound of the self-positioning estimation is derived in closed form.It was shown that our proposed channel estimation and self-positioning method have satisfactory performance,especially when the number of the existing signals increases,the signal-to-noise ratio arises,and the number of the samples in-creases. |
PDF Full Text Request