Tesi etd-06172018-184717 |
Link copiato negli appunti
Tipo di tesi
Tesi di laurea magistrale
Autore
GIANNETTI, FRANCESCO
URN
etd-06172018-184717
Titolo
Strategies and benefits of multi-layer routing
Dipartimento
INGEGNERIA DELL'INFORMAZIONE
Corso di studi
INGEGNERIA DELLE TELECOMUNICAZIONI
Relatori
relatore Prof. Garroppo, Rosario Giuseppe
relatore Prof. Giordano, Stefano
relatore Prof. Giordano, Stefano
Parole chiave
- Multi-layer
- SDN
Data inizio appello
19/07/2018
Consultabilità
Non consultabile
Data di rilascio
19/07/2088
Riassunto
In this paper we focus on dynamically allocating resources to service requests (SRs), which arrive and leave over time and are generated by applications with heterogeneous requirements, in a joint multi-layer and application-aware fashion.
To do so, we use a novel resource allocation algorithm, the Application-Aware (AA) algorithm, able to jointly allocate resources on IP/MPLS and optical layers, while meeting multiple optional application requirements like bandwidth, latency, availability, encryption.
The algorithm exploits the concept of Auxiliary Graphs (AGs), to provide application-awareness for each SR. Moreover, the algorithm has a modular design, i.e., it can be easily extended to consider further application requirements (ARs).
We ran various simulations to evaluate the AA Algorithm performance by comparing them with Baseline (application-unaware) algorithm performance.
We evaluate this two algorithm in terms of both Blocking Probability (BP) and of AR Violation Probability (VP) for the SRs and Convergence Times.
To do so, we use a novel resource allocation algorithm, the Application-Aware (AA) algorithm, able to jointly allocate resources on IP/MPLS and optical layers, while meeting multiple optional application requirements like bandwidth, latency, availability, encryption.
The algorithm exploits the concept of Auxiliary Graphs (AGs), to provide application-awareness for each SR. Moreover, the algorithm has a modular design, i.e., it can be easily extended to consider further application requirements (ARs).
We ran various simulations to evaluate the AA Algorithm performance by comparing them with Baseline (application-unaware) algorithm performance.
We evaluate this two algorithm in terms of both Blocking Probability (BP) and of AR Violation Probability (VP) for the SRs and Convergence Times.
File
Nome file | Dimensione |
---|---|
Tesi non consultabile. |