COMPUTER NETWORKS RESEARCH LAB

TSP Lab

Department of Electrical and Computer Engineering, McGill University

  NAVIGATION

Home
People
Photos
 

  RESEARCH

Projects
Publications
 

  LINKS

AAPN
MITACS
McGill TSP
McGill ECE
 

  LOCAL ACCESS
Local Info
 
 

Project Descriptions Return to AAPN Projects

Online Routing in MPLS Networks
 

Student: Fariba Heidari, Ph.D. Student
Supervisor: Prof. Lorne G. Mason

Description: With the ongoing growth in both the number and the diversity of applications using the Internet, overprovisioning of network resources currently used in IP networks cannot be considered as a suitable long term approach. The application of routing algorithms with the capability of traffic engineering is one of the possible solutions for improving the network resource utilization while satisfying requirements of different applications. The routing algorithms currently used in IP networks are destination-based methods. These algorithms do not update their routing policy based on the current status of the network and current traffic demand.

In this project, a set of event dependent routing schemes with the application to explicit source routing in MPLS networks are presented. The proposed algorithms are based on load shared sequential routing where load sharing factors are updated using reinforcement learning techniques.

The proposed algorithms are simple to implement with low computational cost. These algorithms update their routing policy using their locally observed information and considerably reduce the flooding information overhead to the network. Discrete event simulation results show that the proposed algorithms compare favorably, in terms of blocking probability performance, with alternate event dependent and state dependent routing schemes also proposed for routing in MPLS networks.

One of the issues with most of the state dependent and event dependent routing schemes reported in the literature is that there is no theoretical approach for evaluating their blocking probability performance. In this project, the load share optimization problem in the load shared sequential routing model both from user equilibrium and system optimum perspectives is studied and an analytical method for deriving load shared factors is presented.

Related Publications

F. Heidari, S. Mannor and L.G. Mason, Reinforcement learning-based load shared sequential routing, in Proc. IFIP Networking, Atlanta, GA, May 2007.

G. Brunet, F. Heidari and L.G. Mason, Load shared sequential routing in MPLS networks: system and user optimal solutions, in Proc. EuroFGI NET-COOP, Avignon, France, June 2007.