Access, Resource Allocation, and Performance Analysis in Next Generation Wireless Networks
出版項
2019
說明
1 online resource (231 pages)
文字
text
無媒介
computer
成冊
online resource
附註
Source: Dissertations Abstracts International, Volume: 81-02, Section: B
Advisor: Erkip, Elza
Thesis (Ph.D.)--New York University Tandon School of Engineering, 2019
Includes bibliographical references
With the advent of new technologies and data hungry applications, there has been a substantial increase in the demand for high data rate and low latency in cellular systems. Several architectures and techniques such as Massive multi-input multi-output (Massive MIMO), non-orthogonal multiple access (NOMA), full-duplex (FD) communications, and communications over millimeter-wave (mmWave) bands have been proposed to satisfy these demands. While these techniques can potentially lead to higher data rates and lower delays compared to conventional long-term evolution (LTE) systems, there are various challenges to enable these technologies. The main objective of this dissertation is to investigate a number of these challenges and provide novel and efficient solutions to address them.In NOMA, multiple users are activated in uplink or downlink at the same resource block of the system to enhance resource utilization and improve the system and users throughputs. Similarly, using a FD radio as the base station enables simultaneous uplink and downlink transmissions at the same resource block. A major challenge in NOMA and FD systems is user scheduling and resource allocation which greatly impact the performance in such systems. In the first part of this dissertation, we provide a general framework for opportunistic multi-user scheduling which can be applied to NOMA and FD systems. The objective is to devise scheduling algorithms maximizing the average system utility while satisfying a particular fairness criterion such as temporal fairness or utilitarian fairness. Temporal fair scheduling leads to communication systems with predictable latency whereas utilitarian fairness can be used to model asymmetric traffic demands for users in the system. For each fairness criterion, we show that a particular class of threshold based strategies achieves the optimal average system utility. In these strategies, a fixed threshold is assigned to each user and a subset of users are activated at each resource block based on the users' thresholds as well as channel state information. Furthermore, we propose an online iterative algorithm for each fairness criterion to construct the optimal strategy by finding the optimal thresholds for a given system utility metric. These algorithms do not require knowledge of the users' channel statistics. Rather, at each time-slot, they have access to the channel realizations in the previous and current time-slots.Massive MIMO systems provide several advantages over conventional systems such as higher spectral efficiency and energy efficiency. In the second part of this dissertation, we study Massive MIMO cellular systems operating based on time-division duplexing (TDD) and frequency-division duplexing (FDD). Non-cooperative TDD Massive MIMO, combined with max-min power control, is known to result in significant improvements in per-user throughput compared with conventional LTE technology. We investigate further refinements to TDD Massive MIMO, first, in the form of three-fold sectorization, and second, three-fold sectorization combined with coordinated multi-point operation, in which the three sectors cooperate in the joint service of their users. For these scenarios, we analyze the downlink performance for both maximum-ratio and zero-forcing precoding and derive closed-form lower-bound expressions on the achievable rate of the users. These expressions are then used to formulate power optimization problems with two throughput fairness criteria: i) network-wide max-min fairness, and ii) per-cell max-min fairness. Furthermore, we provide centralized and decentralized power control strategies to optimize the transmit powers in the network.While TDD Massive MIMO can lead to significant throughput multiplexing gains, a large fraction of the cellular market is based on FDD. As opposed to TDD, uplink and downlink channels are not equivalent with FDD. This intensely increases the channel estimation overhead in FDD systems as the time required to estimate the downlink channels is proportional to the number of antennas at the base station. To reduce this overhead while partially exploiting throughput multiplexing gains, we propose several beamforming optimization algorithms based on the long-term channel state information. Using an adaptive scheme, beamforming vectors are updated whenever the long-term channel information changes. First, we study the problem when the base station has a single RF chain (single-beam). Semi-definite relaxation (SDR) with randomization is used to solve the problem in this case. As a second approach, we propose a low-complexity heuristic beam composition algorithm which performs very close to the upper-bound obtained by SDR. Next, we study the problem for a generic number of RF chains (multi-beam) where the Gradient Projection method is used to reduce interference and obtain local solutions.While lower frequency bands are over-crowded, there are large chunks of non-occupied spectrum in mmWave frequencies between 30 GHz and 300 GHz, which can be used to enhance data rates in the next generation wireless networks. In order to cope with high path loss and severe shadowing in mmWave frequencies, it is essential to employ massive antenna arrays and generate narrow transmission patterns (beams). Hence, it is crucial to develop protocols for initiating the communication between a base station and a user with directional beams. Furthermore, mobile user tracking is indispensable for reliable communication when narrow beams are used. In the final part of this dissertation, we investigate efficient beam search and robust user tracking methods for mmWave wireless systems. First, the problem of joint beam search and data communication is studied for static users in a time slotted system where each time slot can be used for beam search (location probing) or data communication. We use a backward dynamic programming approach to find the optimum allocation between probing and data transmission and the beamwidth for location probing that maximize the expected throughput. Next, we propose a joint beam tracking and data communication strategy in which, the base station increases the beamwidth during data transmission to compensate for location uncertainty caused by user mobility. To evade low beamforming gains due to widening the beamwidth, we propose a probing scheme in which the base station transmits a number of probing packets to refine the estimation of angle of arrival based on the user feedback. Furthermore, we provide a steady state analysis based on which, the duration of data transmission and probing phases are optimized
Electronic reproduction. Ann Arbor, Mich. : ProQuest, 2020
