Conference paper Open Access

Optimizing QoE and Cost in a 3D Immersive Media Platform: A Reinforcement Learning Approach

Panagiotis Athanasoulis; Emmanouil Christakis; Konstantinos Konstantoudakis; Petros Drakoulis; Stamatia Rizou; Avi Weit; Alexandros Doumanoglou; Nikolaos Zioulis; Dimitrios Zarpalas


Citation Style Language JSON Export

{
  "publisher": "Zenodo", 
  "DOI": "10.5281/zenodo.3761854", 
  "language": "eng", 
  "title": "Optimizing QoE and Cost in a 3D Immersive Media Platform: A Reinforcement Learning Approach", 
  "issued": {
    "date-parts": [
      [
        2020, 
        4, 
        22
      ]
    ]
  }, 
  "abstract": "<p>Recent advances in media-related technologies, including&nbsp;capturing and processing, have facilitated novel forms&nbsp;of 3D media content, increasing the degree of user immersion.&nbsp;In order to ensure these technologies can readily support the&nbsp;rising demand for more captivating entertainment, both the&nbsp;production and delivery mechanisms should be transformed to&nbsp;support the application of media or network-related optimizations&nbsp;and refinements on-the-fly. Network peculiarities deriving&nbsp;from geographic and other factors make it difficult for a greedy&nbsp;or a supervised machine learning algorithm to successfully foresee&nbsp;the need for reconfiguration of the content production or delivery&nbsp;procedures. For these reasons, Reinforcement Learning (RL)&nbsp;approaches have lately gained popularity as partial information&nbsp;on the environment is enough for an algorithm to begin its&nbsp;training and converge to an optimal policy. The contribution&nbsp;of this work is a Cognitive Network Optimizer (CNO) in the&nbsp;form of an RL agent, designed to perform corrective actions on&nbsp;both the production and consumption ends of an immersive 3D&nbsp;media platform, depending on a collection of real-time monitoring&nbsp;parameters, including infrastructure, application-level and quality&nbsp;of experience (QoE) metrics. Our work demonstrates CNO&nbsp;approaches with different foci, i.e., a greedy maximization of&nbsp;the users&rsquo; QoE, a QoE-focused RL approach and a combined&nbsp;QoE-and-Cost RL approach.</p>", 
  "author": [
    {
      "family": "Panagiotis Athanasoulis"
    }, 
    {
      "family": "Emmanouil Christakis"
    }, 
    {
      "family": "Konstantinos Konstantoudakis"
    }, 
    {
      "family": "Petros Drakoulis"
    }, 
    {
      "family": "Stamatia Rizou"
    }, 
    {
      "family": "Avi Weit"
    }, 
    {
      "family": "Alexandros Doumanoglou"
    }, 
    {
      "family": "Nikolaos Zioulis"
    }, 
    {
      "family": "Dimitrios Zarpalas"
    }
  ], 
  "version": "1", 
  "type": "paper-conference", 
  "id": "3761854"
}
74
50
views
downloads
All versions This version
Views 7474
Downloads 5050
Data volume 39.1 MB39.1 MB
Unique views 6363
Unique downloads 4444

Share

Cite as