Cost-effective deployment strategies for the rural FTTH roll-out

Abstract

The main topic of this dissertation revolves around the problem of ru- ral FTTH service roll-out. The problem has become topical worldwide, typically in the countries with rural populations scattered around. Never- theless, the FTTH service penetration rate generally remains low, which still creates room for research and innovation, whereby the planning of FTTH network deployment and roll-out can be advanced. In fact, the rural FTTH roll-out problem falls into category of worst-case scenarios, thereby driving technological progress to an even higher extent. Inadequate future-proof broadband service in rural areas leads to negative consequences, widely known as the Digital Divide problem. Unequal devel- opment of urban and rural areas deepens the isolation of rural areas from the information society. The resolution of the problem is thus of great importance, also beyond purely technical outcomes. In fact, this work evolves beyond technical aspects and includes recommendations for lo- cal governments, policy regulators, investors, utility companies, and other likely contributors. The first half of this work presents the following content. The background to the cutting-edge broadband technologies was presented with regard to the ultimate success of rural fibre roll-out. The discussion focused on one of the most promising future-proof broadband technologies with re- gard to the rural case, i.e. PON or LR-PON technology. A number of deliberations targeted the statement of the problem, research questions and hypotheses, and finally the research methodology. Key aspects of the problem were further explored. The literature review focused on the state-of-the-art means and deployment technologies, in different localities worldwide, and suitable mathematical formulations needed to model and optimise the deployment. Finally, research gaps were identified, and de- tailed research plan was proposed. The increased deployment cost with respect to expected revenue is over- all perceived as the main barrier to FTTH roll-out in rural areas. This provided grounds for the quest towards more cost-effective deployment strategies. The second half of this work, therefore, focuses on researching the cost- effectiveness of deployment strategies for rural FTTH roll-out. The re- search was confined to the examination of the access section of LR-PON architecture. In fact, the deployment of this part can be classified as one of the most challenging and urgent issues in the rural FTTH roll-out problem. The examination of LR-PON roll-out happens top-down, in the following scopes. Firstly, a potential nationwide deployment of ODN section of 1024-way- split LR-PON in Ireland undergoes examination. The analysis is carried out to assess the effect of different optical power splitter arrangements of PON tree-like topology on the cost-effectiveness of the deployment, i.e. the utilisation of PONs, and the total amount of fibre cable length. A heuristic is used, and hierarchical clustering algorithm proposed with a simplified design, i.e. the Euclidean distance is used to serve as the proximity function in the clustering routine. The research findings and numerical results were discussed. As a result, further improvements to maximise the utilisation of PONs and more advanced deployment models were discussed. Secondly, a specialised ILP model devoted to the rural FTTH deployment and the optimisation of the initial setup cost is proposed and evaluated. This model allows to cover a real-world deployment scenario of PON ar- chitecture and includes a number of important details into the design. Therefore, a least cost solution can be approached, also by taking into account actual equipment pricing and real geographical coordinates. Key features of the model include the use of multi-fibre blown cables, low-count power optical splitters, and street-map data set. Evaluation and tests tar- get various rural areas in Ireland. A single test scenario corresponds to an area covered by a cabinet, which refers to a common migration scenario where FTTC represents an intermediate step towards FTTH. It was shown that by employing small optical power splitters near the locations of end- users, the size of feeder fibre cable decreases, thus significantly improving the initial investment fibre cable cost in excess of 25%. Thirdly, the study adds a relevant perspective into the planning. The dynamism of network growth is included through the techno-economic ex- amination of an important variable, i.e. the service take-up rate. This allows to demonstrate how the problem is further exacerbated by the un- certainty associated with end-users take-up rate. The randomness bound up with the subscribers service take-up imposes fluctuation and escalation in the total cost of deployment, and affects the outcome of the quest for cost-effective deployment strategies. This part of work focuses on the following trade-off: (I) achieving optimal long-term deployment cost, but assuming a known expected take-up rate (due to optimal resource utilisation and sharing) versus (II) sacrificing some optimal resource management, in order to reduce short-term network up-front cost, which reduces the risk associated with lower than expected take-up rate. The author believes that achieving objective (II) is generally key to succeed in rural areas. Moreover, building a network with the lowest up-front cost is critical to approach the viability of the network operation. Indeed, while the first approach (I) has been considered for denser areas, where the favourable economics allow for larger margins in take-up rate uncertainty, this work indicates that the second approach (II) implemented through the deployment strategy proposed herein, is generally preferable for rural roll-out. The numerical results show that the up-front cost can be slashed at the early stage of the FTTH network development. The quest led to deep insight into rural FTTH deployment economics, thereby helping to identify these circumstances under which the deploy- ment can be commercially viable. This could not occur, however, without the application of deployment strategies primarily designed to alleviate and deal with the rural FTTH deployment problem. These methods were mathematically formulated to instantiate a suite of clustering algorithms underpinning the cost-effectiveness and automation of the deployment strategies proposed in this work. Overall, this work proved that rural FTTH roll-out case requires special consideration, and novel network de- sign strategies can further mitigate the problem.