Spectrum Management for QoE-aware Dynamic Spectrum Access Networks

The Hong Kong University of Science and Technology
Department of Computer Science and Engineering


PhD Thesis Defence


Title: "Spectrum Management for QoE-aware Dynamic Spectrum Access
Networks"

By

Miss Yanjiao CHEN


Abstract

The growing demand for wireless communications poses great challenges for 
the current wireless network infrastructure. The risk of spectrum crutch, 
where the user demand outstrips the available spectrum, motivates the 
research on dynamic access to improve the usage of the under-utilized 
spectrum. In addition, video streaming has dominated the wireless traffic, 
and it is important to understand video Quality-of-Experience to 
efficiently improve video service quality over wireless network.

In this thesis, we first study the spectrum management for dynamic 
spectrum access networks from three aspects: spectrum auction, spectrum 
allocation between licensed and unlicensed use, and hybrid access in 
femtocell networks. We identify the challenges under each scenario, and 
propose solutions to address the problem. Then, we conduct a measurement 
study to analyze video Quality-of-Experience, represented by 
Quality-of-Engagement (QoE), in wireless network based on a real-world 
data set.

For the spectrum auction, we propose two mechanisms. First, we design a 
double auction mechanism for heterogeneous spectrums, dealing with the 
problem of heterogeneous interference graphs and heterogeneous buyers' 
valuations. Second, we design an online double auction mechanism for 
dynamic spectrum access, taking into consideration the location-determined 
interference relationship among buyers. For the spectrum allocation 
between licensed and unlicensed use, we build a three-stage game model to 
characterize the interaction among the spectrum regulator, the licensed 
and the unlicensed operators, and the end users. The target is a balance 
of allocation income for the spectrum regulator and social welfare for the 
end users. For the hybrid access in femtocell networks, we establish a 
utility-aware refunding framework, in which the macrocell operator refunds 
the femtocell owners for sharing their femtocell resources.  All of the 
above studies achieve a more efficient utilization of the spectrum 
resources in the dynamic spectrum access networks.

For the measurement study on video QoE over the wireless network, we 
deploy an experimental platform to collect video session data and conduct 
a survey to solicit users' subjective opinions. We first analyze the 
influence of each factor on user QoE, then build a QoE prediction model 
based on regression tree.


Date:			Friday, 9 January 2015

Time:			11:00am - 1:00pm

Venue:			Room 3494
 			Lifts 25/26

Chairman:		Prof. Shiqing Ling (MATH)

Committee Members:	Prof. Qian Zhang (Supervisor)
 			Prof. Kai Chen
 			Prof. Lei Chen
 			Prof. Yu-Hsing Wang (CIVL)
 			Prof. Ming-Hua Chen (Inf. Engg., CUHK)


**** ALL are Welcome ****