Model Transformation for Automatic Design of GPON/FTTH Network

In the literature, automatic design and planning of Gigabit Passive Optical Network/Fiber-To-The-Home (GPON/FTTH) have been addressed by many different approaches including meta-heuristics and Mixed Integer Linear Programming (MILP). The FTTH network design can be formulated as a Binary Integer Programming (BIP) problem with Boolean variables. However, many well-defined optimization approaches for BIP problems developed in the literature cannot be directly applied in network planning due to the complexity of planning model. To address this, the model of GPON planning has been transformed with a standard structure, and the resultant model can be solved efficiently based on the Lagrangian Heuristic method. In addition, the correlations of constraints are considered during optimization process. This helps to reduce the computational cost in different cases considered in the experiments.