Abstract:With the rapid growth of 4G mobile network and vehicular network services，mobile terminal users have increasing demand on data sharing among different radio remote units (RRUs) and roadside units (RSUs). Meanwhile, commercial video-streaming, video/voice conference applications delivered through peer-to-peer (P2P) technology are still keep on stimulating the sharp increment of bandwidth demand in both business and residential subscribers. However, a significant issue is that, although wavelength division multiplexing (WDM) and orthogonal frequency division multiplexing (OFDM) technology have been proposed to fulfil the ever-increasing bandwidth demand in access network, the bandwidth of optical fiber is not unlimited due to the restriction of optical component properties and modulation/demodulation technology, and blindly increase the wavelength cannot meet the cost-sensitive characteristic of the access network. In this paper, we propose a software defined multi-OLT PON architecture to support efficient scheduling of access network traffic. By introducing software defined networking technology and wavelength selective switch into TWDM PON system in central office, multiple OLTs can be considered as a bandwidth resource pool and support flexible traffic allocation for optical network units (ONUs). Moreover, under the configuration of the control plane, ONUs have the capability of changing affiliation between different OLTs under different traffic situations, thus the inter-OLT traffic can be localized and the data exchange pressure of the core network can be released. Considering this architecture is designed to be maximum following the TWDM PON specification, the existing optical distribution network (ODN) investment can be saved and conventional EPON/GPON equipment can be compatible with the proposed architecture. What’s more, based on this architecture, we propose a dynamic wavelength scheduling algorithm, which can be deployed as an application on control plane and achieve effective scheduling OLT wavelength resources between different OLTs based on various traffic situation. Simulation results show that, by using the scheduling algorithm, network traffic between different OLTs can be optimized effectively, and the wavelength utilization of the multi-OLT system can be improved due to the flexible wavelength scheduling.

Cost-efficient Routing, Modulation, Wavelength and Port Assignment Using Reinforcement Learning in Optical Transport Networks

Multi-Objective Routing and Resource Allocation Based on Reinforcement Learning in Optical Transport Networks

A Routing Optimization Method for Software-Defined Optical Transport Networks Based on Ensembles and Reinforcement Learning

Overview on Routing and Resource Allocation Based Machine Learning in Optical Networks

Demonstration of Image Processing Based on Reinforcement Learning in Multi-Modal Optical Transport Networks

Routing, Modulation Level, Spectrum and Transceiver Assignment in Elastic Optical Networks.

Joint Scheduling and Routing for QoS Guaranteed Packet Transmission in Energy Efficient Reconfigurable WDM Mesh Networks

Energy-Efficient Routing, Modulation and Spectrum Allocation in Elastic Optical Networks

Reinforcement Learning (Rl)-Based Holistic Routing and Wavelength Assignment in Optical Network-on-Chip (onoc): Distributed or Centralized?

Leveraging double-agent-based deep reinforcement learning to global optimization of elastic optical networks with enhanced survivability.

Opti-DeepRoute: A Topology-Adaptive Deep Reinforcement Learning Based Service Provisioning Framework for Elastic Optical Network

Optimal Scheduling Model of WDM/OTN Network Transmission Line Based on Machine Learning

DeepRMSA: A Deep Reinforcement Learning Framework for Routing, Modulation and Spectrum Assignment in Elastic Optical Networks

Service Restoration in Multi-Modal Optical Transport Networks with Reinforcement Learning.

Reinforcement-Learning based routing for packet-optical networks with hybrid telemetry

Self-learning Routing for Optical Networks.

RWA Optimization of CDC-ROADMs Based Network with Limited OSNR.

Actor-Critic-Based Resource Allocation for Multi-Modal Optical Networks

Deep Reinforcement Learning-Based RMSA Policy Distillation for Elastic Optical Networks

Integrated Routing In Wavelength-Routed All-Optical Networks

Software Defined Multi-Olt Passive Optical Network for Flexible Traffic Allocation