Provisioning, Recovery, and In-Operation Planning in Elastic Optical Networks

15 Toward Cognitive In‐operation Planning

Alba Pérez Vela¹, Filippo Cugini², Marc Ruiz¹ and Luis Velasco¹

¹ Universitat Politècnica de Catalunya, Barcelona, Spain

² CNIT, Pisa, Italy

Data analytics is an interesting and challenging task to find patterns in heterogeneous data coming from different sources. In networking, data comes from monitoring the data plane, for example, received power, impairments, and errors in the optical layer or service traffic in the Multi‐Protocol Label Switching (MPLS) layer. The output of data analytics can be used for automating network operation and reconfiguration (e.g., see Chapter 11), detecting traffic anomalies, or transmission degradation. This way of doing network management is collectively known as the observe–analyze–act (OAA) loop since it links together monitoring, data analytics, and operation (with or without reoptimization). In this chapter, the OAA loop is applied to a variety of use cases that result into a proposed architecture based on the Application‐Based Network Operation (ABNO) one to support the OAA loop in Elastic Optical Networks (EONs).

Table of Contents and Tracks’ Itineraries

Average	Theoretical	Practitioner
15.1 Introduction
15.2 Data Analytics for Failure Localization
15.2.1 Algorithm for Failure ...