Ad Hoc Cloudlet Based Cooperative Cloud Gaming

Abstract

As the game industry matures, processing complex game logics in a timely manner is no longer an insurmountable problem. Many researchers are now trying to find ways to optimize the gaming system regarding the network usage, local resource utilization, and energy consumption. However, current cloud-based mobile gaming solutions are limited by their relatively high requirements on Internet resources. Also, they typically do not consider the geographical locations of nearby mobile users and thus ignore the potential cooperation among users. Therefore, inspired by existing cloud computing techniques and the concept of ad-hoc cloudlet computing, in this paper, we propose an ad-hoc cloudlet based gaming architecture. Two modules of the architecture are introduced: 1) progressive game resources download, by which mobile users can adaptively download gaming resources from cloud servers or nearby mobile users according to the gaming progress, and 2) ad-hoc cloudlet-based cooperative task allocation, by which gaming components can be executed dynamically over local devices, nearby devices, or cloud servers. We also formulate the mechanism for both modules as an optimization problem and propose several algorithms for both modules, which are later used for evaluation purposes. We carry out simulations based on real mobility traces, and the results show that our system's performance depends highly on the ad-hoc network environment (the more concurrent and balanced connections within the ad-hoc network, the lower the energy costs). Also, regardless of the network environment, our system has lower energy costs while utilizing resources of nearby devices, compared to the cloud-based gaming architecture.