Load balancing using cell range expansion in LTE advanced heterogeneous networks

The use of heterogeneous networks is on the increase, fueled by consumer demand for more data. The main objective of heterogeneous networks is to increase capacity. They offer solutions for efficient use of spectrum, load balancing and improvement of cell edge coverage amongst others. However, these solutions have inherent challenges such as inter-cell interference and poor mobility management. In heterogeneous networks there is transmit power disparity between macro cell and pico cell tiers, which causes load imbalance between the tiers. Due to the conventional user-cell association strategy, whereby users associate to a base station with the strongest received signal strength, few users associate to small cells compared to macro cells. To counter the effects of transmit power disparity, cell range expansion is used instead of the conventional strategy. The focus of our work is on load balancing using cell range expansion (CRE) and network utility optimization techniques to ensure fair sharing of load in a macro and pico cell LTE Advanced heterogeneous network. The aim is to investigate how to use an adaptive cell range expansion bias to optimize Pico cell coverage for load balancing. Reviewed literature points out several approaches to solve the load balancing problem in heterogeneous networks, which include, cell range expansion and utility function optimization. Then, we use cell range expansion, and logarithmic utility functions to design a load balancing algorithm. In the algorithm, user and base station associations are optimized by adapting CRE bias to pico base station load status. A price update mechanism based on a suboptimal solution of a network utility optimization problem is used to adapt the CRE bias. The price is derived from the load status of each pico base station. The performance of the algorithm was evaluated by means of an LTE MATLAB toolbox. Simulations were conducted according to 3GPP and ITU guidelines for modelling heterogeneous networks and propagation environment respectively. Compared to a static CRE configuration, the algorithm achieved more fairness in load distribution. Further, it achieved a better trade-off between cell edge and cell centre user throughputs. [Please note: this thesis file has been deferred until December 2016]

Grantor dc:publisher.institution

Department of Electrical Engineering

Year dc:date.issued

2016

Author dc:creator

• Nyembe, Wiseman Nkosingiphile

Advisor dc:contributor.advisor

• Dlodlo, Mqhele E

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/11427/20338

OAI identifier oai:identifier

oai:open.uct.ac.za:11427/20338