Welcome!

 

This page was created in order to serve as a simple and rapid way to disseminate my work. Here you will find technical reports, summaries of projects and my publications. It also has space for news. In "Alisson's Shared Items" the main news of my feed.

Currently I'm dedicating myself to my dissertation project. The subject is "Traffic Grooming in Optical WDM Networks". Below the goals and motivation of the project.

Traffic Grooming in WDM Networks

May, 2010

While inter-domain routing in the Internet is a widely researched topic, issues concerning inter-domain at the layers below network layer are relatively new. The efficient establishment of connections with high bandwidth between multiples domains separated by technology, administrative rules and different control signals is an open challenge.

In the field of optical networks, a question that has been widely debated, but little has been seen in multi-domain environments, is the traffic grooming in WDM networks. Solutions for traffic grooming in multi-domain scenarios require greater complexity than the intra-domain solutions, given the need to consider several aspects such as signaling overhead, knowledge and communication between the Path Computation Elements, connection set up waiting time, policies, and others.

Moreover, many new applications that demands high bandwidth during a known period of time are emerging. We can consider the example of on-line broadcasting of any major sporting event, where the user buy a service that require a lot of bandwidth during the transmission of the event. It is expected that such applications continue to grow, including new services where the bandwidth is set by the user when buying the service. New architectures are being developed to allow the definition of on-line terms of Service Level Agreement to be fulfilled by the service provider.

Lastly, although rings and point-to-point links will continue to be used for specific technologies and in legacy networks, we can expect more and more networks to be built as meshes. A service provider whose network is based on a mesh is able to provide value-added features such as on-demand provisioning and rapid protection without the massive over-deployment of network resources that would be needed in simpler topologies.

Considering all these aspects, the proposed project aims to study inter-domain dynamic traffic grooming in mesh WDM networks in the cases when the traffic duration is known.