Characterizing the ODN for a PON using longitudinal power monitoring and simplified coherent receivers

Abstract

As passive optical networks (PONs) evolve to meet rising demands in bandwidth and quality of service, accurately monitoring power profiles and thus characterizing the optical distribution network (ODN) has become critically important. This paper first demonstrates the longitudinal power monitoring model for simplified coherent PONs using the regularized linear least squares method with piecewise linear fitting to monitor the power evolution and detect losses along the fiber link. The well-selected regularization was verified to be effective in solving the PON-specific ill-posed problem and mitigating the effects of noise to give a stable solution for longitudinal power monitoring. The proposed model was first verified for single-loss events in the simulation and experiment. Simulation results indicate a maximum quantification error of 0.04 dB and a localization error of 0.10 km under various single-loss insertion scenarios. In the experiment, the single-loss insertion was identified without prior knowledge of the number of losses, achieving maximum quantification and localization errors of 0.14 dB and 0.25 km, respectively, thus demonstrating high loss detection accuracy. The proposed model was also extended to multiple-loss detection, and the simulation results indicate its potential applicability in characterizing the ODN for simplified coherent PONs in practical deployments.