Dimedioning models Ethernet based aggregation ring networks

OData support
Jakab Tivadar
Department of Networked Systems and Services

Nowadays the Ethernet standard is considered to be the market leading technology among home users and SOHO users. The main reason for its high popularity is its simplicity and the elaborate standard. Home users can create a smal LAN network really easily with this technique. Some of the backbone providers still use ATM or frame relay technologies. The paralell use of multiple services raise compatiblity issues and also isn’t very economical to maintain.

However the implementation of the Ethernet in a backbone network isn’t easy, since it was developed for the user side. It doesn’t support OAM functionality and the protection management isn’t solved. The response time of the system in case of failure is much higher than the time of other methods used today.

Multiple standards have been made in order to eliminate all these problems.

My task was to create ring topology network modells and implement them in a linear programming environment. After the mathematical modell was complete I used LP_Solve to implement the network and define different types of services. Some of the traffic on the network went without protection and some of them went with FRR or 1 plus 1 protection. After the simulations were done I check the results and compared them in order to see how much network resources are used in the different cases. I did different types of test, in some cases there were only a few protected traffic components in other cases all of them were protected with a specific protection method.

The models illustrate the resource requirements of the various security services. The observer can see which traffic passes on which path and how much extra load it puts on the network components. The model is suitable for strategic planning support, wheter it is about developing new aggregation rings, or expanding existing network architectures.


Please sign in to download the files of this thesis.