Optimized Client and Line Hardware for Multiperiod Traffic in Optical Networks With Sliceable Bandwidth-Variable Transponders

The sliceable bandwidth-variable transponder (SBVT) has been recently proposed as a transmission module able to combine the scalability of superchannel-based transmission with the flexibility required for highly meshed traffic. The modular nature of the SBVT implies that a large array of transmission capacity is predeployed per module, which significantly affects the planning and dimensioning of backbone optical networks. These planning methods should also account for the switching modules that are required to support the interconnection between the SBVTs on the line-side and the client signal interfaces. In this paper we study the impact of the switch capacity on the cost and flexibility of SBVT-enabled networks. Specifically, we propose a multiperiod planning simulation based on integer linear programming models to optimize incremental traffic deployments with respect to the cost of SBVTs, switches, and client cards. The results highlight the fact that switch capacity must be carefully balanced in order to avoid overprovisioning the switch capacity beyond the traffic requirements, while at the same time providing enough flexibility in the client-to-line interconnection to do efficient lightpath grooming and therefore reduce client-card-and SBVT-related costs.