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

Abstract

An innovative machine learning-based scheduling solution for improving live UHD video streaming quality in highly dynamic network environments. IEEE Transactions on Broadcasting .

A Machine Learning Resource Allocation Solution to Improve Video Quality in Remote Education

“…Moreover, most of the existing TDP-FDP strategies are unable to react to the changeable networking conditions, leading to the over-provisioning of some traffic classes. However, to deal with these challenges, in [5] a Reinforcement Learning(RL)-based solution is proposed to dynamically prioritize one traffic class (i.e. [p, 1, ..., P ], ∀p ∈ P) each time among others.…”

“…The problem is that once selected a scheduling rule r ∈ R, the multi-objective optimization can be unbalanced in the frequency domain since only a particular QoS objective is addressed. Alongside of the traffic class prioritization, in [5] a particular scheduling rule r ∈ R is decided to be applied at each TTI for all traffic classes. However, since the needs in terms of QoS provisioning may differ from one traffic class to another, then a separate scheduling rule may be required.…”

“…[p, P, 1, ..., p + 1, p − 1]) at each TTI; b) the slave controller selects a particular scheduling rule r p ∈ R for each class from the prioritization sequence according to the available bandwidth. Compared to [5], this approach can offer better scalability options and enhanced heterogeneous QoS provisioning.…”

“…However, to be able to accommodate good online learning for everyone and provide an appropriate level of e-learning for students and teachers, there is a need to go beyond the basic requirements of access to a digital device and stable internet connectivity. Quality in all its facets plays a fundamental role and supporting optimal experience for increasing number of users which exchange larger amounts of rich media content across the existing networks is of paramount importance [5].…”

An Innovative Reinforcement Learning-Based Framework for Quality of Service Provisioning Over Multimedia-Based SDN Environments

“…Consequently, guaranteeing QoS provisioning has become an active field of research especially for applications that require data delivery under certain QoS constraints (e.g., interactive multimedia, web conferencing, gaming, etc.). In this context, new emerging technologies and solutions are being explored to accommodate the high traffic demands, such as: Network Function Virtualisation (NFV) and Software Defined Networks (SDN) [3], [4], Advanced Television Systems Committee (ATSC) 3.0 [5], satellite back-haul [6], Multi-Access Edge Computing [7], Unmanned Aerial Vehicles (UAV) and drones [8], machine learning [9], [10].…”