Energy Efficiency Analysis of Multilayer Networks with Time-Varying Traffic
Paper in proceeding, 2025
Global power consumption is growing each year, with a significant contribution from the ICT sector. Although the number of primary devices, such as ROADMs and routers, is usually constant, the number of active transceivers depends on the routing and allocation policy and can be tuned.
This work demonstrates how dynamic network optimization aided by traffic prediction leads to a 16% power saving, expressed as the number of active transceivers, and a 15% provisioned traffic increase.
This work demonstrates how dynamic network optimization aided by traffic prediction leads to a 16% power saving, expressed as the number of active transceivers, and a 15% provisioned traffic increase.
time-varying traffic
Energy efficiency
machine learning.
multilayer network
Author
Aleksandra Knapinska
Chalmers, Electrical Engineering, Communication, Antennas and Optical Networks
Wrocław University of Science and Technology
Piotr Lechowicz
Wrocław University of Science and Technology
Chalmers, Electrical Engineering, Communication, Antennas and Optical Networks
Salvatore Spadaro
Polytechnic University of Catalonia
Krzysztof Walkowiak
Wrocław University of Science and Technology
International Conference on Transparent Optical Networks
21627339 (ISSN)
9798331597771 (ISBN)
25th International Conference on Transparent Optical Networks (ICTON 2025)
Barcelona, Spain,
Barcelona, Spain,
Areas of Advance
Information and Communication Technology
Subject Categories (SSIF 2025)
Communication Systems
Telecommunications
Computer Systems
DOI
10.1109/ICTON67126.2025.11125245