Abstract

ZR+ pluggables have the potential to transform the optical core landscape, paving the way for new, to our knowledge, network services enabled by a reconfigurable infrastructure. However, reconfigurability brings challenges. One of the challenges among them is the fast and accurate planning tools for networks with stochastic demands. In this paper, we introduce SPRINT: a fast and accurate planning engine for ZR+-enabled hop-by-hop IPoWDM networks. By leveraging a novel, to the best of our knowledge, recurrence-based approach, SPRINT can determine the location and number of ZR+ pluggables needed across an IPoWDM network to meet a target quality of service at a very low computational cost. We compared SPRINT to conventional simulation-based dimensioning techniques in real-world network topologies exhibiting different structural, spatial, and spectral features under different traffic loads. Results show SPRINT is up to four orders of magnitude faster while maintaining a high average accuracy of 90%. This speedup is critical when evaluating a number of envisaged scenarios, while accuracy is key to avoid the high cost of overprovisioning. SPRINT supports both uniform and non-uniform dimensioning strategies, providing flexibility in resource allocation and cost reductions in network planning. By addressing the lack of open fast dimensioning tools for hop-by-hop IPoWDM networks, this work contributes to the advancement of the design of next-generation ZR+-based optical networks.