Resource management for cognitive cloud gaming

In contrary to conventional gaming-on-demand solution, cognitive cloud gaming platform facilitates gaming component migration from the cloud server to the players´ terminal, a novel flexible solution to provide Gaming as a Service. In this work, we model the component-based game and investigate the capacity of intelligent resource management for different optimization targets, including cloud resource minimization and throughput-oriented optimization. Experimental results show that, with the cognitive resource management, cloud system is adaptive to various service requirements, such as increasing the quantity of supported devices and reducing the network throughput of user terminals, while satisfying players´ quality of experience.