A novel algorithm for PON optimal deployment over real city maps and large number of users

In this paper we propose a novel algorithm (which combines an integer linear programming ILP model with heuristic approaches) for finding optimal multiple PON deployments in large urban regions with very high number of users and with different bit rate demands. Our algorithm finds optimal solutions employing real city maps (by an ad-hoc developed interface with the open source OpenStreetMap database) using the geographic coordinates of streets and buildings. This data is used for clustering users, identifying their location and demands, allocating primary and secondary street cabinets and finding the optimal routes for optical fiber ducts/cables.