Stochastic QoE-Aware Optimization in Cloud-Based Content Delivery Networks

The problem of cloud resource optimization is examined, while the uncertainty in demand and user feedback is considered. We propose a Markov decision process model for resource assignment in cloud-based content delivery networks. Furthermore, we include a feedback-based probabilistic model for quali...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE access 2018-01, Vol.6, p.32662-32672
Hauptverfasser: Haghighi, Ali A., Shahbazpanahi, Shahram, Shah Heydari, Shahram
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The problem of cloud resource optimization is examined, while the uncertainty in demand and user feedback is considered. We propose a Markov decision process model for resource assignment in cloud-based content delivery networks. Furthermore, we include a feedback-based probabilistic model for quality of experience in the resource assignment problem. We apply dynamic programming to solve this stochastic optimization problem. In order to address the challenges regarding the computational complexity of the problem, we first present an optimal solution with linear complexity for a special case of unlimited bandwidth cloud sites. Then, we propose a sub-optimal algorithm for the generic bandwidth-constrained problem with significantly reduced complexity and quasi-optimal performance. Simulation results are presented to corroborate the merits of the proposed algorithms.
ISSN:2169-3536
2169-3536
DOI:10.1109/ACCESS.2018.2845740