Enhancing Video Streaming Quality of DASH over Cloud/Edge Integrated Networks

Abstract

With the advancement of mobile technologies and the popularity of mobile devices, mobile video streaming applications/services have increased considerably in recent years. Dynamic Adaptive Streaming over HTTP (DASH) or MPEG-DASH is one of the most widely used video streaming techniques over the Internet. It adapts video sending bit rate according to available network resources, however, in case of low bandwidth, DASH performs poorly, which will cause video quality degradation and video stalling.

Mobile Edge Computing (MEC) or Multi-access Edge Computing, in connection with the backend cloud has been used to reduce latency and overcome some of the video quality degradation problems for mobile video streaming services. However, an end user might be suffering from video quality drop downs when s/he moves out from the coverage of one node to another or when the mobile network condition goes down. To tackle the degradation problems and assure enhanced video streaming quality, a novel follow-me Edge Node Prefetching (ENP) scheme was proposed and developed in the project, by prefetching video segments in advance in the upcoming node used by the end-user. A test bed was set up consisting of a backend cloud (OpenStack), two edge nodes (LXD Containers) and one mobile device, the ENP algorithm was implemented on the cloud server and client sides. Experiments were carried out for the DASH streaming service based on Dash.js from the DASH Industry Forum. Preliminary results show that the ENP scheme can maintain higher video quality and less service migration time when moving from one mobile node to another, when compared to existing approaches, i.e. live migration in Follow-me-Edge and the C-up schemes. The proposed scheme could be useful in smart city applications or providing seamless mobile video streaming services in Cloud/Edge integrated networks.