Date of Award

12-2019

Document Type

Dissertation - Campus Access Only

Degree Name

Doctor of Philosophy (PhD)

Degree Discipline

Electrical Engineering

Abstract

The proliferation of smart mobile devices and applications requiring high data rates has resulted in phenomenal mobile data growth around the world, with no growth in the scarce spectral resources. The demand for these scarce spectral resources will increase with the introduction of fifth generation (5G) wireless communication systems. Considering the 5G use cases, ranging from the Broadband Access, Internet of Things (IoT), Public Safety and Autonomous Driving, leading to a multi-tiered heterogeneous wireless architecture, further strain on these limited resources is expected. 5G is being designed to operate across a vast range of frequency bands, spanning licensed, shared, and unlicensed spectrum. Therefore, an intelligent and efficient use of scarce spectrum is highly essential. This study examines performance gain from applying machine learning to smart radio terminals, resulting in better spectral efficiency. Specifically, the dissertation examines application of machine learning to spectrum awareness using Spectrum Data (Radio Frequency (RF) traces) to improve spectral efficiency. A machine learning based spectrum awareness model trained using historical spectrum data can learn and make accurate prediction of a complicated coexistence scenarios in heterogeneous networks, such as deciding if the spectrum is idle (available for use) or whether to coexist with spectrum occupant or seek for other spectrum if number of users will degrade performance or a licensed user is present. Traditional spectrum awareness method can only decide whether the spectrum is idle or busy, this will not ensure the spectral efficiency we seek in future wireless networks due to missed transmission opportunities. Future work will focus on application of novel machine learning models to address complexity challenges in future wireless communication resource allocation optimization such as vehicular communication, exploiting other wireless big data such as user mobility data and network management data.

Committee Chair/Advisor

Lijun Qian

Committee Member

Xiangfang Li

Committee Member

John Fuller

Committee Member

Pamela Obiomon

Committee Member

Matthew Sadiku, Zhu Han

Publisher

Prairie View A&M University

Rights

© 2021 Prairie View A & M University

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

Date of Digitization

11/25/2024

Contributing Institution

John B Coleman Library

City of Publication

Prairie View

MIME Type

Application/PDF

Download

Share

COinS