Enabling Superior and Controllable Video Streaming QoE with 5G Network Orchestration

5G promises to offer a single network platform that can be optimized for several vertical industries and applications. It is envisioned that this optimization will be achieved through advanced network orchestration functionalities. Video streaming accounts for the vast majority of mobile traffic, hence optimizing performance and Quality of Experience (QoE) for video traffic becomes crucial for the next generation (5G) mobile networks. In this paper, we focus on video streaming applications based on MPEG-DASH (Dynamic Adaptive Streaming over HTTP), a technology that dynamically adapts to changing network conditions by requesting content in chunks encoded at different bitrates, and which is used by major streaming services like Netflix and YouTube. The adaptation process of DASH clients may be speeded up by using the recently standardized MPEG-DASH SAND (Server and Network-assisted DASH) protocol, which envisions the presence of a DANE (DASH-Aware Network Element) in the network providing real-time network bandwidth information to DASH clients. However, when DASH clients do not request the content at the bitrate recommended by the DANE, i.e. they ignore SAND messages or purposely request content at higher bitrates, bandwidth may end up being used in suboptimal way and end users will likely end up receiving poor playback experience, especially on shared bottlenecks. In this paper, we address this issue by combining SAND with network orchestration, where we use advanced SDN (Software Defined Networking) capabilities to have the network enforce the bandwidths selected by the DANE for each individual DASH client, and we demonstrate this concept in the state of the art 5G testbed of the H2020 TRIANGLE project.