Deep Learning-Based Dynamic Bandwidth Allocation for Future Optical Access Networks

Abstract

Over the last decade, Passive Optical Networks (PONs) have emerged as an ideal candidate for next-generation broadband access networks. Meanwhile, machine learning and more specifically deep learning has been regarded as a star technology for solving complex classification and prediction problems. Recent advances in hardware and cloud technologies offer all the necessary capabilities for employing deep learning to enhance Next-Generation Ethernet PON's (NG-EPON) performance. In NG-EPON systems, control messages are exchanged in every cycle between the optical line terminal and optical network units to enable dynamic bandwidth allocation (DBA) in the upstream direction. In this paper, we propose a novel DBA approach that employs deep learning to predict the bandwidth demand of end-users so that the control overhead due to the request-grant mechanism in NG-EPON is reduced, thereby increasing the bandwidth utilization. The extensive simulations highlight the merits of the new DBA approach and offer insights for this new line of research.

Keywords

• Deep learning
• dynamic bandwidth allocation
• machine learning
• NG-EPON
• optimization
• PON