1. UKM Learning and Research Repository
2. PHD Thesis / Tesis PHD
3. Faculty of Engineering and Built Environment / Fakulti Kejuruteraan dan Alam Bina

Title:	Semi-virtual clustering based co-tier interference management techniques in ultra-dense femtocell networks
Authors:	Ibrahim Ali M. Shgluof (P57621)
Supervisor:	Mahamod Ismail, Prof. Dr.
Keywords:	Femtocells
Issue Date:	1-Feb-2018
Description:	Heterogeneous networks comprising both macro and femtocells have been one of the main proposals adopted by the Third Generation Partnership Project (3GPP) for Long Term Evolution-Advanced systems to extend the coverage area and enhance network capacity. However, cross-tier interference is one of the main issues in such deployment scenarios, particularly when femtocells share the same channel. Dedicated channel assignments for femtocells may resolve this issue, but leads to underutilization of resources and thus; degrades the capacity of the system. Nonetheless, with the immense growth of traffic volume in recent years due to the emergence of smart devices, the 3GPP have shifted their focus towards the technology of small cell densification. In this regard, femtocell densification seems to be the most favorable solution in the future Fifth Generation (5G) networks, specifically for traffic concentrated in hot zones. Nevertheless, femtocell densification may lead to, among others, severe co-tier interference, especially when femtocells operate in Closed Subscriber Group mode. On this subject, several femtocell clustering techniques for interference management have been proposed. However, these approaches do not distinguish between users within the same femtocell with regard to interference levels, hence undermining the resource utilization within the victim cells and the whole network. The main objective of this research is to propose semi-virtual clustering based co-tier interference management techniques in ultra-dense femtocell networks. Firstly, the optimal location of a small femtocell cluster under macro cell area is identified for improving indoor coverage and enhancing user's throughput, while allowing co-channel deployments. Secondly, a dynamic femtocell classification technique been proposed to categorize femtocells into different Classes: Victim, victim-aggressor, aggressor and neutral, based on the status of their respective users. Thirdly, an adaptive semi-virtual clustering based co-tier interference management algorithm (SCVC) been proposed; without the use of an initiative fixed frequency arrangement. This algorithm virtually semi-cluster each victim femtocell depending on the status of its active users. Then, a disjoint cluster is built for each victim cell and its aggressor/s. Moreover, a dynamic mechanism for optimal resource allocation is implemented independently within each disjoint cluster. This mechanism allows the use of frequency reuse one for non-critical users, while the critical users have to share the resources with their aggressors. The extensive network simulation results concern mainly among others, the mean throughput of critical users, the average capacity of victim femtocells and the percentage of resource utilization, with two approaches of QoS requirements; unconstrained and constrained. For the former approach, the proposed SCVC algorithm outperformed one of the prominent schemes, namely Femtocell Cluster-Based Resource Allocation scheme in all different deployment densities; that range from sparse to ultra-dense by an approximate average gain of 185%, 64% and 31% for the aforementioned three metrics respectively. These results are with respect to the state where no coordination occurs between femtocells and all share the same resources at the same time. In future work, the proposed techniques can be rehearsed and tested for applicability to mitigate co-tier interference and optimize radio resources in the upcoming 5G ultra-dense networks.,Ph.D.
Pages:	162
Call Number:	TK5105.78.S535 2018 3 tesis
Publisher:	UKM, Bangi
Appears in Collections:	Faculty of Engineering and Built Environment / Fakulti Kejuruteraan dan Alam Bina

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.