MACHINE LEARNING-BASED MODEL FOR QOE PREDICTION IN CLOUD-BASED MULTIPLAYER GAMES USING NETWORK QOS PARAMETERS

Abstract

Cloud-based multiplayer gaming has become an application domain where latency is crucial to user experience in cloud-gaming environments and network performance will determine how well players enjoy their games. Variations in these Quality-of-Service parameters such as packet loss, jitter and latency create gameplay responsiveness, stability and user satisfaction degradation that are directly impacted by these variables. Machine-Learning algorithms have been developed to predict the Quality-of -Experience based on key network Quality-of-Service metrics within cloud-based multiplayer gaming environments. A dataset was created through systematic changes made to the levels of jitter, latency and packet loss and corresponding scores for the quality of experience were derived from standardized evaluation metrics. Multiple machine-learning algorithms used to model the relationship between impairment in network Quality-of-Service and Quality-of-Experience which enables precise estimation under different conditions. Performance evaluation demonstrated that proposed model captures effectively the impacts of Quality-of-Service parameters and has high accuracy for predictions As compared to the traditional methods. Proposed framework delivers a scalable-solution for real-time quality of experience estimation and supports adaptive optimization of network resources in cloud-gaming environments.