Cost-Effective Network Planning and Operation for Rural Communities.

Abstract

Broadband Internet access is central to the regeneration of remote communities and reducing the digital divide between rural and urban regions. This thesis focuses on rural communities with limited financial resources, environmental issues including long reach from conurbations, and mountainous or otherwise adverse terrain, typically with limited access to a wired power supply. As such, regular access technologies based on cable or fibre optics are not financially viable. To overcome this challenge, we consider the deployment of a Free-Space Optical (FSO) based relay network as the primary technology, using diversity to provide resilience to atmospheric effects. The aim of this research is to design and evaluate a rural network planning and traffic engineering framework employing FSO communication using light emitting diodes/lasers to construct backhaul rural network infrastructures. FSO systems are relatively cheap and easy to implement [1]. Various proof-of-concept technologies already exist [2] [3] [4]. However, the focus of this work is on the design of a flexible network-planning tool together with a robust management framework that is designed to operate over such an infrastructure to ensure it functions efficiently despite changes in load or communication channel outages. Although the work concentrates on an FSO based infrastructure, this could be extended to support heterogeneous networks employing a combination of technologies. More precisely, this research first describes a novel network planning tool with an intelligent resource management system based on a Multi-Objective Evolutionary Algorithm (MOEA) that determines the suitable location of FSO relay nodes, taking into account end-to-end link speed which is bitrate of user data and the degree of path diversity coupling with battery power. This MOEA approach can account for Line-of-Sight occlusions and allows various compromises to be selected from a Pareto front to suit individual needs. We provide suitable results to show the satisfactory operation of the tool and outline avenues for future development. Following on from this, we design and evaluate an intelligent traffic-engineering framework to make the best use of the deployed infrastructure that can adapt to environmental changes. This aims to ensure a good service is maintained at all times by suitable reconfiguration.