Two distributed reactive MPLS traffic engineering mechanisms for throughput optimization in best effort MPLS networks

This paper describes a scalable distributed reactive MPLS traffic engineering system, which can be used for throughput optimization of best effort traffic in IP networks. The load rebalancing in the system is realized either by path rerouting or by multi path balancing. To prevent the system from routing oscillations, the distributed traffic engineering units coordinate the load rebalancing actions by using routing update messages. It is shown by simulation, that the reactive MPLS traffic engineering system performs well in great networks and optimizes the network throughput dramatically compared to shortest path routing. The path rerouting and multi path balancing approach reach comparable results while the multi path balancing performs slightly better in the presented scenario. To evaluate the reactive MPLS traffic engineering approaches in realistic network scenarios, a scalable rate based simulation environment is used.