Year
2023
Degree Name
Master of Philosophy
Department
School of Electrical, Computer and Telecommunications Engineering
Abstract
The deployment of extremely large antenna arrays (ELAAs) at transceivers is a promising technology for next-generation wireless communication. With the increase in the number of antennas at the receiver and the adoption of high-frequency communication, the transmitters and/or scatterers may be located within the near- eld region of the receiver, constituting near- eld communication. As a result, traditional far- eld techniques may become inapplicable, and the channel's near- eld nature demands innovative solutions. Furthermore, ELAA systems have substantial power needs, particularly when fully digital architectures are used. To address this challenge, hybrid architectures are considered as a viable alternative. However, during each channel use, the number of observations is limited at hybrid receivers. Therefore, near- eld channel estimation in ELAA systems requires increased overhead and computational complexity.
This thesis investigates near- eld localization and channel estimation for ELAA systems. A hybrid receiver equipped with a uniform linear array (ULA) is considered in a single-input multiple-output (SIMO) con guration. We propose a subarraybased solution that reduces the antenna aperture to e ectively shrink the near- eld range. By doing so, the transmitters and/or scatterers are positioned within the far eld of the subarray, thereby enabling the application of far- eld-based techniques. Firstly, we estimate subarray directions of arrival (DoAs) and times of arrival (ToAs) by using atomic norm minimization (ANM) and root multiple signal classi cation (root-MUSIC) techniques. Secondly, the transmitter/scatterers' locations are jointly estimated from the multiple subarray DoAs and the corresponding subarray ToAs after association. Then, to improve the channel estimation performance, the estimated locations are re ned using grid search or gradient descent. Finally, the near- eld channel is reconstructed based on the re ned transmitter/scatterers' locations. Subsequently, we also propose a generalized nested sampling (GNS)-based alternative to reduce the computational complexity. Simulation results demonstrate that the proposed solutions can achieve high-accuracy localization and channel estimation. Besides, the complexity of the proposed estimator can be alleviated with the proposed GNS-based method, as only a selected number of receiver antennas need to be activated.
Recommended Citation
Chen, Xuan, Subarray-Based Near-Field Channel Estimation for Extremely Large Antenna Array Systems, Master of Philosophy thesis, School of Electrical, Computer and Telecommunications Engineering, University of Wollongong, 2023. https://ro.uow.edu.au/theses1/1851
FoR codes (2008)
0906 ELECTRICAL AND ELECTRONIC ENGINEERING
Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.