Abstract

Modern broadband hybrid optical-wireless access networks have gained the attention of academia and industry due to their strategic advantages (cost-efficiency, huge bandwidth, flexibility, and mobility). At the same time, the proliferation of Software Defined Networking (SDN) enables the efficient reconfiguration of the underlying network components dynamically using SDN controllers. Hence, effective traffic-aware schemes are feasible in dynamically determining suitable configuration parameters for advancing the network performance. To this end, a novel machine learning mechanism is proposed for an SDN-enabled hybrid optical-wireless network. The proposed architecture consists of a 10-gigabit-capable passive optical network (XG-PON) in the network backhaul and multiple Long Term Evolution (LTE) radio access networks in the fronthaul. The proposed mechanism receives traffic-aware knowledge from the SDN controllers and applies an adjustment on the uplink-downlink configuration in the LTE radio communication. This traffic-aware mechanism is capable of determining the most suitable configuration based on the traffic dynamics in the whole hybrid network. The introduced scheme is evaluated in a realistic environment using real traffic traces such as Voice over IP (VoIP), real-time video, and streaming video. According to the obtained numerical results, the proposed mechanism offers significant improvements in the network performance in terms of latency and jitter.

Highlights

  • Modern telecommunication networks become increasingly complicated with various features and potentials

  • When x-axis is 10, the network begins with a User equipment (UE) per Optical Network Units (ONUs)-evolved Node B (eNB) during the first minute, it continues

  • The proposed learning approach is based on knowledge obtained by Software Defined Networking (SDN) controllers regarding the uplink and downlink traffic requests of the mobile end-users

Read more

Summary

Introduction

Modern telecommunication networks become increasingly complicated with various features and potentials. Flexible network architectures, which are capable of adapting different vertical adaptation requirements, are feasible by industry players and operators [4, 5] Demanding applications such as Voice over IP (VoIP), Video on Demand (VoD), High-Definition Television (HDTV), and online gaming entail a high-capacity network infrastructure that would allow mobility, flexibility, and Quality of Service (QoS) provisioning. Hybrid optical-wireless networks integrate (a) broadband wireless solutions that are able to support multiple, and even mobile, end-users, such as Worldwide Interoperability for Microwave Access (WiMAX), third generation partnership project (3GPP), Long-Term Evolution (LTE), and Wireless Fidelity (Wi-Fi), for example, legacy IEEE 802.11, in a mesh environment and (b) optical technologies that bring huge bandwidth, in specific, fixed, and predefined optical paths, such as the passive optical network (PON) technology [6].

Motivation and Contributions
Related Work
Architecture
Machine Learning Mechanism
Performance Evaluation
Conclusions
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call