Abstract
Due to the emergence of numerous bandwidth-hungry applications, we are motivated to investigate cheaper and faster Internet access solutions to serve in a neighborhood. We concentrate on the convergence of optical and wireless networks for the deployment of Internet access networks so that we can exploit the opportunities of both technologies. We focus on network dimensioning and placement of equipment in hybrid optical-wireless access networks. A number of integrated optical-wireless architectures have been investigated for the greenfield deployment of future access networks. A novel hybrid network infrastructure, namely PON-LTE access network, has been proposed where fiber will be deployed as deeply as affordable/practical and then, wireless systems will be used to extend this connectivity to a large number of locations and ultimately connect the wire-less end users. We propose a 2-phase network design optimization scheme for greenfield deployment of PON-LTE access network infrastructure. In the first phase, we propose an integer linear programming (ILP) model which optimizes the greenfield deployment of LTE network based on the static distribution of mobile user equipment (MUE). The proposed model takes into account various physical layer constraints of LTE network and determines the optimal clustering of MUEs as well as the location of eNBs in a neighborhood. In the second phase, we propose another ILP model which determines the optimal placement of a number of integrated ONU-EPCs as well as a single AWG with respect to the location of the eNBs determined during the first phase. Our proposed scheme also facilitates the construction of an efficient physical architecture/topology of a hybrid PON-LTE access network. Computational experiments have been conducted on three different data sets consisting of 128, 256 and 512 mobile user equipment in order to evaluate the performance of the proposed scheme.