IP and Transport networks are controlled and operated independently today, leading to significant Capex and Opex inefficiencies for the providers. We discuss a unified approach with OpenFlow, and present a recent demonstration of a unified control plane for OpenFlow enabled IP/Ethernet and TDM switched networks. IntroductionWide area networks are expensive to own from a service provider perspective and it is widely understood that much of this cost is in operational expenses. However, service providers such as AT&T and Verizon are obliged to own and operate two distinct networks for IP and Transport. These networks are typically planned, designed and managed by separate divisions within the same organization, leading to substantial management overhead, functionality/resource duplication, and increased Opex. This issue has been widely discussed over the last decade.An important aspect of these networks is that the two do not dynamically interact. The IP network does not exploit the full capabilities offered by the underlying Transport network because IP routers are typically interconnected via static circuits created in the Transport network. The IP network simply views these circuits as fixed pipes that make up the (virtual) IP links. The natural consequence is that carriers do not benefit from dynamic circuit switching in the core, where the latter can offer significant advantages. For example, circuit switches are more easily scalable, switch at higher data rates and consume much less power than packet switches. A useful rule of thumb is that an optical circuit switch consumes a 10 th of the volume and power, and costs 10 times less than an equivalent electronic packet switch with the same capacity. While circuit switches lack the statistical multiplexing benefits of packet switches, this is of less significance in the core, where flows are naturally aggregated and relatively smooth. If the network operations could be unified, and circuits could be dynamically created, modified and destroyed, the service provider could benefit from a more cost-efficient and energy efficient converged network. Such a network would scale better to rising traffic and changing service needs, while reaping the benefits of both kinds of switching technologies -packet switching and dynamic circuit switching.While we are not the first to suggest the benefits of such convergence, we believe one of the main reasons preventing such convergence is the control and management architecture of the two networks. IP networks have control mechanisms that are fully-distributed and tightly linked to the task of packet forwarding in each switch and router in the network. Because of the tight linkage, it becomes difficult to introduce new routing mechanisms or applications into the network. Contrastingly, Transport networks have traditionally had a clean separation between control and data planes, but have no visibility into IP traffic patterns and application requirements, and have been managed primarily through EMS/NMS systems under the ...

No abstract

This document defines an architecture framework and functional requirements for transport of signaling information over IP.

No abstract

No abstract