National Technical University of Athens (NTUA)
Αλγόριθμοι δρομολόγησης για ευέλικτα οπτικά δίκτυα
Abstract
dc:descriptionhis PhD dissertation belongs to the research field of optical networks, and more specifically to the field of routing and optimization algorithms for elastic (flexible) optical networks. The continuous Internet traffic increase has created the need for more efficient optical networks. Recent developments in the field of optical networks are particularly important but are not enough to meet the needs of the near future. The architecture of the optical networks should evolve to allow dynamic control and adaptation of the network to changing conditions. This presupposes the development of algorithms that will initially use the information from the current state of the network. Next, from current and future requirements, the algorithms will determine possible optimization functions, and then decide on the execution of the respective actions. The implementation of these algorithms is the subject of this doctoral dissertation and has the effect of increasing the efficiency of optical networks, managing their resources more efficiently and reducing operating costs. In particular, we developed three methods for estimating the quality of transmission. The first is based on an algebraic method (Network Kriging) that models the interference of neighboring channels to provide high accuracy in the estimation. The other two estimation methods use mechanical learning and achieve even better accuracy. One uses mechanical learning methods to train the parameters of a physical level model to achieve high accuracy, and the other uses machine learning techniques with specific features to train to provide high-precision assessments. We also developed a framework that also relies on Network Kriging and reduces the metering inaccuracies of performance monitors, and detects link-level errors. Furthermore, we developed a toolkit that is used to adjust the transmission parameters of the paths (symbol rate, modulation format etc) so that their quality of transmission is always acceptable. Finally, we developed an ILP algorithm that enables spectrum sharing between different class connections. In case the quality of transmission of a high class path deteriorates, it can borrow spectrum from a low class neighbor and adjust its transmission parameters to bring the quality of transmission back to acceptable levels.
Degree
thesis:*- Grantor dc:publisher
- National Technical University of Athens (NTUA)
- Year dc:date
- 2019
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Sartzetakis, Ippokratis-Leonidas
- Σαρτζετάκης, Ιπποκράτης-Λεωνίδας
Subjects
dc:subject × 20- Ελαστικά οπτικά δίκτυα
- Βελτιστοποίηση δικτύων
- Μηχανική μάθηση
- Δρομολόγηση και ανάθεση φάσματος
- Εκτίμηση ποιότητας μετάδοσης
- Επιβίωση δικτύου
- Εγγυήσεις ρυθμού
- Διαστρωματική βελτιστοποίηση
- Elastic optical networks
- Network optimization
- Machine learning
- Routing and spectrum allocation
- Quality of Τransmission (QoT) stimation
- Network survivability
- Rate-guarantees
- Cross layer optimization
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Engineering and Technology
- Electrical Engineering, Electronic Engineering, Information Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/46027