EFFECTIVE RESOURCE ALLOCATION IN HOME EVOLVED NODE BASE STATION NETWORKS

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


PhD Thesis Defence


Title: "EFFECTIVE RESOURCE ALLOCATION IN HOME EVOLVED NODE BASE STATION 
NETWORKS"

By

Miss Ying WANG


Abstract

Thanks to their ability to enhance transmission quality by offloading LTE 
Evolved Node Base Station (eNB) traffic, Home eNB (HeNB) deployment has 
progressed dramatically in recent years. Despite this popularity, the 
deployment of HeNBs has also introduced a new set of resource inefficiency 
problems, caused mainly by inter-cell interference and indoor traffic load 
fluctuation. A large body of previous work has been devoted to improving the 
utilization of network resources in HeNBs, including, frequency planning, time 
division, power control, and space division. Nevertheless, many problems not 
addressed adequately by these studies remain open, in particular those related 
to communication overhead, computational complexity, and service quality.

In the absence of standardized resource allocation mechanisms for LTE HeNBs, in 
this work we focus on designing effective solutions to the resource allocation 
problem in HeNB networks; and address three operational environments: i) 
open-access enterprise networks, ii) closed-access residential networks, and 
iii) open-access residential networks.

First, we focus on improving the resource utilization in centrally-controlled 
open-access enterprise networks such as enterprise HeNB networks andWireless 
Local Area Networks (WLAN). To the best of our knowledge, our work is the first 
to jointly optimize user association, antenna beam selection, link scheduling, 
and power adaptation in such networks. We proposed a unified conflict-free 
scheduling algorithm, that can be directly implemented in HeNB networks, and 
also designed a coordination protocol, the so-called TD-CSMA protocol to enable 
deployment over uncoordinated WLAN MAC protocol.

Next, identifying the difficulty of implementing X-2 interfaces among HeNBs, we 
propose fully distributed resource allocation solutions for non-cooperative 
distributed closed-access residential networks. We introduced a self-learning 
HeNB MAC protocol to mitigate interference by relying on historical user 
feedback. The learning process is modelled as a cluster-structural regret-based 
learning game, where the users within one HeNB form a cluster to share 
information.

Finally, we examined the problem of the energy cost of running a massive number 
of always-on HeNBs worldwide and focused on designing an energy efficient, 
QoS-constrained MAC protocol for cooperative distributed open-access 
residential HeNB networks. Since most previous work has failed to properly 
consider interference mitigation when designing such protocols, we manipulate 
user association and OFDMA scheduling with a combination of interference 
mitigation measures. We proposed two iterative algorithms guaranteed to 
converge based on inter-HeNB communication.


Date:			Tuesday, 2 August 2016

Time:			3:00pm – 5:00pm

Venue:			Room 5564
 			Lifts 27/28

Chairman:		Prof. Shing Yu Leung (MATH)

Committee Members:	Prof. Brahim Bensaou (Supervisor)
 			Prof. Gary Chan
 			Prof. Kai Chen
 			Prof. Wai-Ho Mow (ECE)
 			Prof. Ashfaq Khokhar (Elec. & Comp. Engg.,
 					      Illinois Institute of Tech)


**** ALL are Welcome ****