Context-aware HTTP Adaptive Streaming in mobile cloud environments

Abstract

With advances of cloud computing, seamless video streaming from video server to cloud client has been one of technical challenges for multimedia cloud applications. Especially in case that Desktop-as-a-Service (DaaS) as a major cloud application is deployed via wireless networks, it could raise a new set of issues to be addressed. To solve the problem, we propose a Cloud-based Context-aware HTTP Adaptive Streaming (C2HAS) agent located at cloud server. The goal of the agent is to maximize the video quality of seamless streaming perceived by cloud client, given a dynamically changing network context. From network context we derive a major metric for adapting and maximizing the video quality perceived by cloud clients, which is the throughput ratio of backbone networks and access networks. Based on the metric, we can provide a maximal quality of seamless video streaming to cloud users who might be connected via distant and/or lossy wireless links. The experimental performance analysis shows that the C2HAS agent could be a viable solution for cloud-based multimedia applications.