Improving Carrier Ethernet Recovery Time Using a Fast Reroute Mechanism

Ethernet has evolved from a local network technology to a technology that can be used in metro access and transport networks. Unfortunately the new Ethernet networks cannot achieve the level of reliability of traditional TDM Carrier-class networks. In order to improve Ethernet performance, we need to develop new mechanisms specifically adapted to this type of network technology. The high convergence time of control protocols is one of the problems in both traditional and new generation Ethernet networks. This paper presents a fast recovery mechanism as a solution for the high convergence time. This mechanism uses tunnels and cycles to provide a local recovery and is adapted to a Carrier-class Ethernet network controlled by a link-state protocol. Simulations are used to show the advantage of the use of this mechanism in comparison to the global recovery mechanism of link-state protocols.