19th Biennial Conference of the International Telecommunications Society (ITS): "Moving Forward with Future Technologies: Opening a Platform for All", Bangkok, Thailand, 18th-21th November 2012
The two most significant factors that affect the deployment of Next Generation Access (NGA) networks are the cost of the investment and the expected demand for the new fibre-based services. The related literature is based on very simplified assumptions regarding cost and demand structures. In particular, the investment cost is assumed to be increasing and convex reflecting the fact that fibre deployment becomes marginally more expensive as it is extended to rural, less populated areas. In addition, the demand for the new fibre-based services is estimated by assuming that a certain level of NGA investment leads all consumers to equally increase their willingness to pay for such services. This article contributes to the emerging research on the investment in access infrastructures. In particular, the assumptions about cost and demand structures are modified in order to capture the access networks' underlying morphology complexity and the consumers' socioeconomic characteristics, respectively. Firstly, an empirical analysis is conducted for the 100 major municipal departments from urban to rural in Greece. Their street network data are analyzed as the basis of the NGA installation combining GIS technology and Graph Theory techniques and hence the main cost-drivers are derived. Using regression analysis a real-data-based cost function is obtained. Secondly, a novel model that takes into account socioeconomic characteristics affecting the impact of a certain level of NGA investment on consumers' willingness to pay is developed. The Pareto consumer distribution is used to reflect the greater (lower) positive impact of NGA investments on the willingness to pay of the consumers who live in more (less) populated areas. The comparison of the existing models with the ones developed in this paper shows that: (i) the cost function used in the existing models always underestimates the investment cost of the higher populated areas and overestimates the investment cost of the lower populated areas; (ii) the demand for the new fibre-based services is higher under the proposed than the existing approach; and (iii) the level of NGA investment chosen by the investor is always much higher under the proposed than the existing approach.