STAR-HUB VS. OADM ARCHITECTURES IN LARGE METRO AREA NETWORKS

摘要

Fiber is the most cost-effective way to move huge amounts of bandwidth over short and long distances. In this paper, we have evaluated the metropolitan area network by establishing cost curves comparing the approaches of leasing and building fiber networks. In a typical metro environment, building breaks even with leasing at around 20 buildings served. Having established a baseline for fiber ownership, we then examined the differences between a shared Optical Add/Drop Multiplexed (OADM) ring and a star-hub topology. The star network was shown to have a lower electronics cost in small networks, where the cost savings come primarily from obviating the need for amplifiers. That cost advantage proved true for the star as the network grew to 20 buildings. As the number of buildings on the network increases, star networks scale linearly, while the need to cross-connect different multi-node rings causes OADM ring networks to incur operational complexity in the form of ring interconnection at a central hubbing site. We demonstrated the star network is simpler to deploy, and, as the network grows, easier to groom on more wavelengths and add additional buildings. Finally, it was noted that LGN's strategy of entering a market as a carrier's carrier while deploying networks with a large footprint is best supported by a star network. By owning high fiber count networks with 20-25 buildings per metro, LGN can capitalize on the economies of scale of the star topology, thereby providing maximum bandwidth, flexibility and scalability to its customers. The table below summarizes these findings.