Title: A smart network and compute-aware Orchestrator to enhance QoS on cloud-based multimedia services

Authors: Rodrigo Moreira; Flávio De Oliveira Silva; Pedro Frosi Rosa; Rui L. Aguiar

Addresses: Faculty of Computing, Federal University of Uberlândia, Uberlândia, Minas Gerais, Brazil ' Faculty of Computing, Federal University of Uberlândia, Uberlândia, Minas Gerais, Brazil ' Faculty of Computing, Federal University of Uberlândia, Uberlândia, Minas Gerais, Brazil ' Telecommunications Institute, University of Aveiro, Aveiro, Portugal

Abstract: Rich-media applications deployed on cloud lead the use of the internet by people and organisations around the world. The advent of Software-Defined Networking (SDN), Networking Function Virtualisation (NFV), and Machine Learning (ML) techniques brings new possibilities to address carrier environment challenges, especially QoS. The literature does not show a smart and flexible solution that brings scalability with holistic management of network, compute, and application resources taking into account different approaches to enhance QoS. This paper presents the Orchestrator that manages agnostic network, compute, and applications resources hosted in the cloud. Through ML algorithms, our solution enhances QoS towards network resilience, bandwidth allocation based on real-time traffic, and end-to-end QoS mechanism to the event-driven scenario. The Orchestrator operation spans across different domains providing applications, virtual functions, and cloud resources management elastically. Our experimental evaluation in a large-scale testbed shows Orchestrator's capability to provide a smart jitter decrease using AI techniques.

Keywords: SDN; software-defined networking; network function virtualisation; QoS; machine learning; cloud computing.

DOI: 10.1504/IJGUC.2020.103969

International Journal of Grid and Utility Computing, 2020 Vol.11 No.1, pp.49 - 61

Received: 07 Sep 2018
Accepted: 26 Oct 2018
Published online: 04 Dec 2019 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article