Geometric versus geographic models for the estimation of an FTTH deployment

Abstract

Optical access networks provide a future proof platform for a wide range of services, and today, several operators are deploying fibre to the home (FTTH) networks. Installing an FTTH infrastructure, however, involves very high investment cost. Therefore, a good estimation of the investment cost is important for building a successful business strategy and, consequently, to speed up the FTTH penetration. In this paper, for calculating the amount of cable and fibre in the outside plant together with the associated civil works, and the number of required network elements, two different approaches are investigated: (1) geometric modelling of the fibre plant based on approximate mathematical models and (2) geographic modelling of the fibre plant based on map-based geospatial data. The results obtained from these two approaches can then be used as input for preliminary investment cost calculations and/or techno-economic evaluations. Compared to more complex and accurate geographic modelling, we verify that especially with uneven population density and irregular street system, simple geometric models do not provide accurate results. However, if no geospatial data is available or a fast calculation is desired for a first estimation, geometric models definitely have their relevance. Based on the case studies presented in this paper, we propose some important guidelines to improve the accuracy of the geometric models by eliminating their main distortion factors.

Dimensioning and Assessment of Protected Converged Optical Access Networks

“…The working FF is the shortest fiber path, whereas the protection is the longer one. The ring is computed using the Travelling Salesman Problem (TSP) [6]. The TSP available at ArcGIS is based on a tabu search-based algorithm to find the best sequence of visiting the stops by preserving the first and last node (either AWG or MCO).…”

“…In practice, the areas where optical access networks are deployed are not evenly populated and the fiber trenching is constrained by various local conditions, e.g., parks, rivers, railways or highways. This is a reason why the geometric models cannot contribute to the accurate estimation of the deployment cost [6].  Geographic models are the most preferred by operators, because of their high accuracy which ensures getting realistic results.…”

A Green Open Access Optical Distribution Network with Incremental Deployment Support

“…Business Scenario: The business scenario is depicted in Table III and is discussed next. We assume that users are distributed across a Manhattan simplified street length (SSL) geometric model [25] with a span of 2 × 2 km 2 for calculating the fiber and trenching distances. In our business scenario, the infrastructure is designed to support a maximum of 16384 (2 14 ) subscribers.…”

Cost evaluation of innovative offers using detailed equipment, process and network modeling languages

“…1b indicates the distribution section from the street cabinet to individual houses, which will typically follow the sidewalks and which is analytically approached as a uniform double street length model. The graphical user interface for PNMN allows the user to choose the deployment model (not limited to the choices presented above, see [3]), and indicate the relevant parameters: the population density and the surface in km 2 . The model then provides the trenching (or aerial deployment) distance, as well as the total cable and fiber length.…”