Quality of experience driven control of interactive media stream parameters

In recent years, cloud computing has led to many new kinds of services. One of these popular services is cloud gaming, which provides the entire game experience to the users remotely from a server, but also other applications are provided in a similar manner. In this paper we focus on the option to render the application in the cloud, thereby delivering the graphical output of the application to the user as a video stream. In more general terms, an interactive media stream is set up over the network between the user's device and the cloud server. The main issue with this approach is situated at the network, that currently gives little guarantees on the quality of service in terms of parameters such as available bandwidth, latency or packet loss. However, for interactive media stream cases, the user is merely interested in the perceived quality, regardless of the underlaying network situation. In this paper, we present an adaptive control mechanism that optimizes the quality of experience for the use case of a race game, by trading off visual quality against frame rate in function of the available bandwidth. Practical experiments verify that QoE driven adaptation leads to improved user experience compared to systems solely taking network characteristics into account