Cost-Effective and Reliable Multi-Period Planning for Fixed and Flexible Grid Architectures in Optical Networks

This work studies cost-effective and reliable multi-period planning in transparent optical networks using the C-band, considering two different architectures: fixed and flexible grid. We propose two Integer Linear Programming (ILP) formulations for minimum-cost network planning, guaranteeing 1 + 1 protection for every demand. The two-stage Sequential ILP optimizes the planning of the working paths of the demands first and then optimizes their protection. The second ILP, referred to as Joint ILP, performs the planning simultaneously for both working and backup paths for every demand. The ILPs minimize the planning costs based on a cost model that considers the equipment cost of establishing a new lightpath and lighting a dark fiber when and where it is necessary for the network. Results showed that over a 4-year planning period, the two methods provided similar costs both for the fixed and the flexible grid architectures but presented a slight difference in the required fibers. For the fixed grid, the Joint ILP needs one more fiber than the Sequential ILP, while for the flexible grid, the number of fibers was the same for both approaches. Overall, the number of required fibers was 5 – 7 times lower in the flexible grid compared to the fixed grid, but the costs were 25–55 % higher for the flexible grid. However, with the expected upcoming maturity of the flexible grid components, the costs reduce to comparable levels with the fixed grid after 2 – 3 years.