Monte Carlo based channel estimation and evaluation for 5G millimeter-wave systems

Current wireless communications systems utilizing frequencies below 6 GHz are quickly becoming unable to meet consumer needs due to limited bandwidth. The need for ever increasing data throughput for wireless systems is pushing systems to begin using higher millimeter-wave frequencies to provide frequency bandwidths that are orders of magnitude larger than those operating at the lower frequencies. Operating at millimeter-wave frequencies brings along new challenges that do not exist at lower frequencies such as mass produced hardware that is physically smaller to overcome the larger free space loss which can bring increase uncertainty in channels and hardware components. The uncertainty increases in these systems must be corrected by channel estimators to correctly receive data. Channel estimators are e↵ected by the increased hardware error and their efficacy is therefore reduced. This in turn increases the error in received data and reduces the throughput of a communications system. A multitude of channel estimation techniques have been developed in an attempt to reduce bit errors in the received data and maximize the system throughput. Methods can be as basic as performing division and averaging of known transmitted values to estimate the channel. Other channel estimators have been designed using techniques ranging from machine learning to optimization and minimization. None of these estimators use responses and uncertainties of hardware components that are typically provided by manufacturers. By directly using known hardware responses and any corresponding uncertainties, a channel estimator can improve its estimation. The research outlined here first provides a framework for the consistent evaluation and comparison of estimators in millimeter-wave systems and second develops an improved channel estimator that leverages known hardware responses and their uncertainties. The comparison between estimators is important to select the best estimator for a scenario before deployment of hardware to save time and money. Without a framework for this comparison, there is no way to directly test channel estimators without implementing and deploying them into a physical system. The developed framework provides common communications metrics in phased array architectures and propagation environments to evaluate how di↵erent channel estimators perform. This framework is then used to evaluate a novel Monte Carlo augmented channel estimator. This channel estimator uses repeat simulations of the known hardware responses and their uncertainties to provide an improved channel estimation. This channel estimator is directly compared to other commonly implemented channel estimators to demonstrate how it can be used in communications systems to reduce overall bit errors in received data and increase system throughput.

Name thesis:degree_name

Doctor of Philosophy (Ph.D.)

Level thesis:degree_level

Doctoral

Discipline thesis:degree_discipline

Electrical Engineering

Grantor dc:publisher

Colorado School of Mines. Arthur Lakes Library

Year dc:date.issued

2022

Author dc:creator

• Weiss, Alec

Advisor dc:contributor.advisor

• Elsherbeni, Atef Z.

Committee members dc:contributor.committeemember

• Quimby, Jeanne
• Hadi, Mohammed
• Aaen, Peter H.
• Fasshauer, Gregory

• channel estimation
• millimeter-wave
• Monte Carlo
• orthogonal frequency division multiplexing
• phased arrays

Statement dc:rights

• Copyright of the original work is retained by the author.

Language dc:language.iso

eng, English

Identifier

T 9372

OAI identifier oai:identifier

oai:repository.mines.edu:11124/15464