QoE-Centric control and management of multimedia services in software defined and virtualized networks

Abstract

Multimedia services consumption has increased tremendously since the deploy­ ment of 4G/LTE networks. Mobile video services (e.g., YouTube and Mobile TV) on smart devices are expected to continue to grow with the emergence and evolution of future networks such as 5G. The end user's demand for services with better quality from service providers has triggered a trend towards Quality of Experience (QoE) - centric network management through efficient utilization of network resources. However, existing network technologies are either unable to adapt to diverse changing network conditions, or limited in available resources. This has posed challenges to service providers for provisioning of QoE-centric multimedia services. New networking solutions such as Software Defined Networking (SDN) and Network Function Virtualization (NFV) can provide better solutions in terms of QoE control and management of multimedia services in emerging and future networks. The features of SDN, such as adaptability, programmability and cost effectiveness make it suitable for bandwidth intensive multimedia applications such as live video streaming, 3D/HD video and video gaming. However, the delivery of multimedia services over SDN/NFV networks to achieve optimized QoE, and the overall QoE-centric network resource management remain an open question especially in the advent development of future softwarized networks. The work in this thesis intends to investigate, design and develop novel ap­ proaches for QoE-centric control and management of multimedia services (with a focus on video streaming services) over software defined and virtualized networks. First, a video quality management scheme based on the traffic intensity under Dynamic Adaptive Video Streaming over HTTP (DASH) using SDN is developed. The proposed scheme can mitigate virtual port queue congestion which may cause buffering or stalling events during video streaming, thus, reducing the video quality. A QoE-driven resource allocation mechanism is designed and developed for improving the end user's QoE for video streaming services. The aim of this approach is to find the best combination of network node functions that can provide an optimized QoE level to end-users through network node cooperation. Furthermore, a novel QoE-centric management scheme is proposed and developed, which utilizes Multipath TCP (MPTCP) and Segment Routing (SR) to enhance QoE for video streaming services over SDN/NFV-based networks. The goal of this strategy is to enable service providers to route network traffic through multiple disjointed bandwidth-satisfying paths and meet specific service QoE guarantees to the end-users. Extensive experiments demonstrated that the proposed schemes in this work improve the video quality significantly compared with the state-of- the-art approaches. The thesis further proposes the path protections and link­ failure free MPTCP/SR-based architecture that increases Survivability, resilience, availability and robustness of future networks. The proposed path protection and dynamic link recovery scheme achieves a minimum time to recover from a failed link and avoids link congestion in softwarized networks.