Dissertation/Thesis Abstract

Spectrum Sharing in Large-Scale and Random Geometric Wireless Networks
by Cai, Ran, Ph.D., The Chinese University of Hong Kong (Hong Kong), 2014, 168; 3691911
Abstract (Summary)

The demand for larger user traffic capacity and better service quality for wireless communications has been increasing drastically in the past decade due to the widespread use of internet and smart phones. However, such demand is severely limited by the shortage of radio spectrum. One of the key enabling techniques to enhance spectrum utilization efficiency is spectrum sharing, which allows unlicensed secondary users to access the spectrum of a licensed primary network under interference constraints. Traditional spectrum sharing strategies developed for finite and deterministic networks require accurate information on user locations or channel gains. It is trivial that acquiring such information consumes substantial resources in large-scale and randomly deployed wireless networks. In this thesis, innovative spectrum sharing techniques for large-scale and random geometric wireless networks are explored by utilizing stochastic geometry models. Our study covers situations where each node of both primary and secondary networks is equipped with one or more than one antenna. We also consider the scenario when multiple secondary networks are present in the communication system.

We first review the various important aspects of spectrum sharing between one primary network and one secondary network where each node is equipped with only one single antenna. We analyze the successful transmission probability of each network in this case, and derive the corresponding optimal transmit power for the secondary network. The main technical challenge is to manage intra-network and inter-network interference caused by the stochastic nature of channel propagation and node distribution. Given a decrement limit for the successful transmission probability of each network, the optimal transmit power of the secondary network is determined to boost the spectrum sharing throughput while preventing individual networks from experiencing severe performance degradation.

Secondly, two multiple-input single-output networks are investigated to further improve the performance of spectrum sharing. Multiple transmit antennas manage aggregate interference by strengthening the desired signal and nulling undesired interferers, as far as possible. Partial zero-forcing beamforming is applied to spectrum sharing networks to quantify the possible density increase in the secondary users while meeting the outage requirements of other spectrum sharing users. Furthermore, we analyze the effects of the nulled interferers, and show how spectrum sharing opportunities can be enhanced in large-scale and random geometric wireless networks by wisely allocating the degrees of freedom for interference nulling.

Lastly, motivated by the evolution of wireless networks toward heterogeneity, we study spectrum sharing between one primary network and multiple secondary networks that are distinguished by system parameters, such as network densities and target data rates. A power allocation strategy is developed for the secondary networks to improve the overall spectrum sharing throughput while guaranteeing the quality-of-service of each network. The joint power allocation problem is transformed into a power ratio allocation strategy, and a quasi-closed form solution that allows for water-filling interpretation is obtained.

Indexing (document details)
Advisor: Ching, Pak-Chung
Commitee:
School: The Chinese University of Hong Kong (Hong Kong)
School Location: Hong Kong
Source: DAI-B 76/08(E), Dissertation Abstracts International
Source Type: DISSERTATION
Subjects: Computer Engineering, Electrical engineering
Keywords: Convex optimization, Multiple antennas, Power controls, Spectrum sharing, Stochastic geometry, Wireless networks
Publication Number: 3691911
ISBN: 9781321668988
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy
ProQuest