An Innovative Machine-Learning-Based Scheduling Solution for Improving Live UHD Video Streaming Quality in Highly Dynamic Network Environments

Abstract

The latest advances in terms of network technologies open up new opportunities for high-end applications, including using the next generation video streaming technologies. As mobile devices become more affordable and powerful, an increasing range of rich media applications could offer a highly realistic and immersive experience to mobile users. However, this comes at the cost of very stringent Quality of Service (QoS) requirements, putting significant pressure on the underlying networks. In order to accommodate these new rich media applications and overcome their associated challenges, this paper proposes an innovative Machine Learning-based scheduling solution which supports increased quality for live omnidirectional (360°) video streaming. The proposed solution is deployed in a highly dynamic Unmanned Aerial Vehicle (UAV)-based environment to support immersive live omnidirectional video streaming to mobile users. The effectiveness of the proposed method is demonstrated through simulations and compared against three state-of-the-art scheduling solutions, such as: static Prioritization (SP), Required Activity Detection Scheduler (RADS) and Frame Level Scheduler (FLS). The results show that the proposed solution outperforms the other schemes involved in terms of PSNR, throughput and packet loss rate.