Cloud Mobile Gaming: Enabling Rich Internet Multi-player Gaming on Mobile Devices

CMG_small.jpg

Overview

The emergence of new and more capable mobile devices along with the steady deployment of broadband wireless networks is making mobile access to rich Internet sites a reality. This technological progress opens up a new possibility: the ability to play rich Internet games produced for PCs on wireline networks from mobile devices. However, due to the inherent hardware constraint of mobile devices such as memory and graphics processing, the goal might be difficult to achieve using the current client-server gaming architecture for PC-based Internet games, where most of the storage and computational burden of the game lies with the client device.

In this work, we study a new cloud server based approach for mobile gaming, termed Cloud Mobile Gaming (CMG), where a cloud gaming server is responsible for executing the appropriate gaming engine and streaming the resulting gaming ideo to the client device, while the mobile devices simply communicate the users' gaming commands to the cloud server. In our work, we first investigate the major challenges for this new CMG approach, including user experience (including game response time), mobile network bandwidth, cloud computing cost, and scalability to a large number of CMG users. Subsequently, we propose and develop several techniques to address the mentioned challenges. We first developed and validated a Mobile Gaming User Experience (MGUE) model, which can quantitatively measure user perceived mobile gaming experience. Next, we develop an adaptive rendering technique which can simultaneously vary the richness and complexity of graphic rendering to adapt the communication and computing needs of each CMG session in responding to the dynamic conditions of the wireless networks and cloud server. Finally, we present a mobile cloud scheduling approach which can allocate resources to meet user experience requirements while maximizing the number of users which can be scheduled concurrently and minimizing cloud cost. 

Associated Publications