PDF of this tutorialThe development of metro optical networks is not a simple, two-step story of rings and then mesh networks but a more gradual evolution. Carriers have been establishing point-to-point optical links to enterprise customers to take advantage of immediate revenue opportunities for the last 18 months. Star and ring topologies will accommodate a boom in network traffic and growing demand for ubiquitous services over the next year. Eventually, carriers will build out full cross-connect, optical-fiber meshes as carrier infrastructures naturally begin to emulate traditional data networks, starting in 2001 and 2002.
Point-to-Point Links
The revenue potential of gigabit-speed services rivals that of traditional leased-line services, but legacy transport systems are not up to the opportunity. Carriers’ most advantageous option is adding high-speed, point-to-point optical links outside their existing core networks. The carrier must deploy only a customer-premises device for converting application traffic to specific wavelengths and multiplexing. Carriers today can extend their local networks as large-customer opportunities arise, and, one day, these island links can be integrated with the core.
Stars/Rings
At the onset of this phase in the development of metro optical networks, carriers are delivering more and higher-speed access services. The traffic burden on their networking infrastructures is mounting. Some carriers increase their transport capacity offers to other carriers, and, thus, fiber exhaustion develops in some parts of their metro networks.
As this phase develops, the creation of star topologies or linear rings will be undertaken gradually, with the simple connection of separate point-to-point links as revenue opportunities dictate. This will be of prime benefit to carriers’ network planners, whose primary consideration must be relying on optical-networking solutions that utilize open network-management systems.
As the number of customers requesting high-speed services explodes and bandwidth-on-demand services begin to appear, carriers with linear-ring topologies will evolve toward full rings based on optical add-drop-multiplexer platforms. Despite some vendors’ suggestions to the contrary, carriers generally have not yet reached this stage, as the number of customers demanding gigabit services is still small.
But, in fact, data traffic is growing at an enormous pace, and it is critical for carriers to build a networking infrastructure that anticipates tomorrow’s realities. With full optical rings, time to market with services will be slashed and carriers will save via reduced labor costs managing their networks.
Configurable protection capabilities will be of prime importance in the services platforms enabling this phase. A carrier will be able to determine whether an individual channel is to be fully protected on the optical layer or entrusted to already-implemented protection mechanisms from the SONET/synchronous digital hierarchy (SDH) layer. As a result, reselling capacity to other carriers and service providers will be made more cost-effective.
Meshes
The day is approaching where the logical infrastructure of glass fiber rings, present in today’s cities, ceases dictating network design. High-end users will seek flexible lambda services, and carriers will need to optimize the optical network for this demand.
With the introduction of metro optical cross-connects, optical metro networks will be transformed into fully flexible platforms. The transformation of the traditionally voice-based network to a data-centric infrastructure will be complete.
The most important technologies in optical-networking solutions for this phase will be cost-effective optical-electrical-optical cross-connects with node-management and network-signaling software functionality similar to that of open shortest path first (OSPF) and multiprotocol label switching (MPLS).
