Feature-based optical spectrum monitoring for failure detection and identification

© 2019 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes,creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other worksIn this paper, we explore the benefits of analysing the optical spectrum of lightpaths for soft-failure detection and identification in Spectrum Switched Optical Network. We present a framework exploiting machine learning (ML) based algorithms that uses descriptive models of the optical spectrum of a lightpath in different points along its route to detect whether the optical signal experiences anomalies reflecting a failure in the intermediate nodes. Our proposal targets the two most common filter-related soft-failures; filter shift (FS) and filter tightening (FT), which noticeably deform the expected shape of the optical spectrum. In this regard, filter cascading is a key challenge as it affects the shape of the optical spectrum similar to FT. Our proposals avoid the misclassification of properly operating signals when normal filter cascading effects is present. Extensive numerical results are presented to compare the performance of the proposed approaches in terms of accuracy and robustness.This work has received funding from the European Commission’s H2020 METRO-HAUL project (G.A. nº 761727), the AEI/FEDER TWINS project (TEC2017-90097-R), and ICREA InstitutionPeer Reviewe