Cloud Mobile 3D Display Gaming User Experience Modeling and Optimization by Asymmetric Graphics Rendering
S. Dey,Yao Liu,Yao Lu
DOI: https://doi.org/10.1109/JSTSP.2015.2396475
2015-01-23
Abstract:With the arrival of auto-stereoscopic 3D displays for mobile devices, and emergence of more 3D content, there is much anticipation for 3D mobile multimedia experiences, including 3D display gaming. Simultaneously, with the emergence of cloud computing, more mobile applications are being developed to take advantage of the elastic cloud resources. In this paper, we explore the possibility of developing Cloud Mobile 3D Display Gaming, where the 3D video rendering and encoding is performed on cloud servers, with the resulting 3D video streamed over wireless networks to mobile devices with 3D displays for a true 3D mobile gaming experience. However, with the significantly higher bitrate requirement for 3D video, ensuring user experience may be a challenge, both in terms of 3D video quality and network delay (response time), considering the bandwidth constraints and fluctuations of wireless networks. In this paper, we propose a new asymmetric graphics rendering approach which can significantly reduce the video encoding bitrate needed for a certain video quality, thereby making it easier to transmit the video over wireless network. However, since asymmetric graphics rendering may also impair the graphics quality, we need to be able to understand and measure its impact. We conduct subjective tests to study and model the impairments due to asymmetric graphics rendering and network delay, thereby developing a user experience model for cloud based mobile 3D display gaming. By conducting subsequent subjective tests, we prove the correctness of the impairment functions and the resulting user experience model. Furthermore, given any network condition, we propose to solve the problem of selecting the optimal graphics rendering factors for the left and right views so as to maximize user experience of cloud mobile 3D display gaming. In order to solve this problem, we first develop a model to estimate the resulting video bitrate of the rendered 3D video when certain graphics rendering factors are used. Next, we derive a model to predict the delay given the available network bandwidth and the video bitrate of the rendered 3D video. We use the above two models together with a branch and bound algorithm to solve the optimization problem and determine the optimal values for the left and right view rendering factors. Experiments conducted using real 4G-LTE network profiles on commercial cloud service demonstrate the feasibility of significant improvement in user experience when the proposed optimization algorithm is used to dynamically select optimal rendering factors according to changing network conditions.
Engineering,Computer Science