Vés al contingut (premeu Retorn)

Doctoral thesis lecture by Hamzeh Khalili

20th September, Hamzeh Khalili will be defending his doctoral thesis entitled "Contributions towards softwaritzation and energy saving in passive optical networks" at 11:30am in room C4-001.
Doctoral thesis lecture by Hamzeh Khalili

Date: 20th September, 2018

Room: C4-001, EETAC (how to arrive)

Author: Hamzeh Khalili, telematic engineer and reseracher at i2CAT Foundation.

Title: Contributions towards softwaritzation and energy saving in passive optical networks.

Directors: David Rincon Rivera and Sebastian Sallent Ribes.

Summary: This thesis is a result of contributions to optimize and improve the network management system and power consumption in Passive Optical Networks (PON). Passive Optical Network elements such as Optical Line Terminal (OLT) and Optical Network Units (ONUs) are currently managed by inflexible legacy network management systems. Software-Defined Networking (SDN) is a new networking paradigm that improves the operation and management of networks by decoupling control plane from data plane.

In this thesis, we propose a novel architecture, based on the SDN concept, for Ethernet Passive Optical Networks (EPON) that includes the Service Interoperability standard (SIEPON). In our proposed architecture, the OLT is partially virtualized and some of its functionalities are allocated to the core network management system, while the OLT itself is replaced by an OpenFlow switch. A new MultiPoint MAC Control (MPMC) sublayer extension based on the OpenFlow protocol is presented. The OpenFlow switch is extended with synchronous ports to retain the time-critical nature of the EPON network. Our simulation based results demonstrate the effectiveness of the new architecture, while retraining a similar (or improved) performance in terms of delay and throughput when compared to legacy PONs. In addition, we introduce an energy-efficient Dynamic Bandwidth Allocation (DBA) algorithm for both the upstream and downstream channels of EPON. The proposed algorithm analyzes the queue status of the ONUs in order to turn ONUs to doze/sleep mode when there is no upstream/downstream traffic in the network, respectively.