top of page

Objectives

Demonstration of AI-assisted network control strategies to proactively identify any soft failure of inline network equipment (amplifiers) and its utilization to reduce network downtime under the C+L Band optical network.

Deliverables

ML-assisted QoT Estimator: Estimates the quality of transmission (QoT) of lightpath with an accuracy of 1 dB margin while considering several input features (such as link length, launch power, etc.) of a light path as input.

ML-assisted QoT Degradation Classifier: Classify the combination of degrading in-line equipment while monitoring the QoT degradation of the lightpath with an accuracy >= 95%.

Network Simulator: Embedding the Classifier and QoT estimator into a state-of-art Multiband (C+L band) network simulator and demonstrating network automation by taking pre-emptive actions to prevent network downtime.

Product Phase of Deliverables: As provisioned by DOT to demonstrate the performance of the classifier over Tejas/C-DOT's testbed.

Co-Principal Investigator:

Anand Srivastava

Professor

Indraprastha Institute of Information Technology (IIIT) Delhi

Principal Investigator:

Abhijit Mitra

Assistant Professor

Indraprastha Institute of Information Technology (IIIT) Delhi

WP14

AI-Enabled Pre-Emptive Network Control in Multiband Optical Networks

Organization:

No. of Members:

2

Member Details:

-

bottom of page