Model Validation of DSCM Transceivers From B2B Characterization for Disaggregated Optical Networks

In this work, we present a predictive model for optical transceivers based on Digital Subcarrier Multiplexing (DSCM), targeting disaggregated optical network scenarios. The model operates at the level of individual subcarriers and is derived from back-to-back measurements, where the transceiver is treated as a closed box. By fitting the relationship between Received Optical Power (ROP), Signal-to-Noise Ratio (SNR), and bit error rate for each subcarrier, the approach captures performance variability across subcarriers and modulation formats. To improve the accuracy of classical formulations, an empirical correction is introduced to better match the observed behavior, especially at high SNRs. This enables a more reliable estimation of performance metrics and reduces prediction error across operating conditions. The resulting model provides a compact representation of the behavior of DSCM-based transceiver that could be exploited, for example, by network control frameworks to estimate performance without continuous monitoring. This work establishes the basis for future integration into digital replicas of optical networks, enabling performance prediction and automated optimization.