Seyed Mahdi Jameii - Department of Computer Engineering, Shahr-e-Qods Branch, Islamic Azad University, Tehran, Iran
Kourosh Khanzadi - Department of Computer Engineering , Science and Research Branch, Islamic Azad University, Tehran, Iran

Unlike traditional gaming where a game run locally on a user's device, in cloud gaming, an online video game runs on remote servers and streams directly to a user's device. This caused players to become independent of having high hardware resources in their local computers. Since video games are a kind of latency-sensitive application, cloud servers far from users are not suitable. In fog computing, fog nodes are in the vicinity of users and can reduce the latency. In this paper, a latency reduction method based on reinforcement learning is proposed to determine which computing fog node can run the video games with the lowest latency. In the proposed method, a Principal Component Analysis (PCA) based approach is used to extract the most important features of each video game as the input of the learning process. The proposed method was implemented using Python. Experimental results show that the proposed method compared to some existing methods can reduce the frame latency and increase the frame rate of video games.

Reinforcement Learning, Cloud-fog gaming, Principal component analysis, Latency Frame rate