There is no irony like the present. It seems that the instant something is declared outdated, it is immediately recycled as the 'Next Big Thing.' The latest technology to rise from the dead is circuit-based networks. The latest incarnation of circuits is now referred to as "lightpaths" or "lambdas" as they typically use optical fiber-based infrastructure, often with multiple light frequencies to provide more than one high-capacity channel. In reality, lightpaths are more than just the fibers, but however you look at them, they are essentially point-to-point connections that are currently wired by hand -- circuits.
Two of the most advanced academic networks in North America have been driving the use of lightpaths (or the related term "lambdas") through similar initiatives. In both cases, the trend to a converged, IP-over-everything future is being challenged by a new vision that isn't all that new in some ways. Today's Internet isn't up to the demands of the future (see "Fix before breaking") and solutions are currently being sought.
Canada's CANARIE has pioneered the User Controlled Light Path (UCLP) that allows end-users (or the fabled intelligent provisioning system) to treat network resources as software objects -- paths can be reconfigured and provisioned on-demand either within a single domain or between/across independently managed domains. The objective is to allow communities requiring high-performance networks to allocate resources, such as end-to-end network connections, as needed without intervention from core network managers.
Several versions of the UCLP software are now available that operate with CA*net4 and support on-demand allocation. The tools are still somewhat crude (e.g. CLI) but effectively demonstrate CANARIE's vision of a dynamically provisioned, de-centralized, high-performance next-generation network. While the majority of the user population will take advantage of the usual routed CA*net4 IP network, certain communities (such as high energy physicists sharing data or bio-chemists involved in automated drug discovery) will be able to "soft-wire" custom network connections between specific end-points, effectively creating a private high-performance network, on either a short- or long-term basis.
In the U.S., the National Lambda Rail (NLR) and Internet2's Hybrid Optical and Packet Infrastructure (HOPI) have been pursuing complementary objectives. The objective of the NLR is to support multiple, contiguous high performance networks on a single core network. Created by and for the U.S. research community as a network technology test bed, the NLR supports a broad range of networks that employ the same optical infrastructure. The HOPI project combines the NLR resources with existing packet-based networks into a single scalable architecture, effectively extending the NLR capabilities into other network infrastructures. In June 2005, NLR and Internet2 announced a merger of their efforts to enhance the value of overall NLR/HOPI effort.
The primary objective of NLR/HOPI is to develop a network infrastructure that is not intrinsically tied to TCP/IP and its inherent limitations. Applications and communities that require custom technologies or transport protocols will be able to use a common core system that also serves the current packet-base TCP/IP networks. Similar to CANARIE project, this means that network resources can be allocated on-demand to meet the needs of the moment.
For both CANARIE and Internet2, the vision for lightpaths includes phrases like "dynamic provisioning" and "automated resource allocation." For example, Internet2 has a research project called Dynamic Resource Allocation over GMPLS Optical Networks (DRAGON) – it will provide a common control plane architecture to support provisioning circuit-oriented services on varied multi-domain networks. However, the early phases of these next-gen networks will be largely limited to manual configuration and the use of static point-to-point paths. Once set up, they remain in place until they are "torn down" again and re-allocated.
So what does this mean today? Circuits. Very, very fast circuits. Soft circuits that can be re-wired by command. Offering test-bed capabilities for alternate, scalable transport technologies. In some cases allocated by end-users instead of operators.
Despite the clear benefits, the irony remains: Just when we thought IP packets would rule the future and networks would finally be comfortably converged, we "discover" that we can't really live without circuits -- at least for now.
Plus ça change, plus ça reste la même chose.
Chief Scientist for Apparent Networks, Loki Jorgenson, PhD, has been active in computation, physics and mathematics, scientific visualization, and simulation for over 18 years. Trained in computational physics at Queen's and McGill universities, he has published in areas as diverse as philosophy, graphics, educational technologies, statistical mechanics, logic and number theory. Also, he acts as Adjunct Professor of Mathematics at Simon Fraser University where he co-founded the Center for Experimental and Constructive Mathematics (CECM). He has headed research in numerous academic projects from high-performance computing to digital publishing, working closely with private sector partners and government. At Apparent Networks Inc., Jorgenson leads network research in high performance, wireless, VoIP and other application performance, typically through practical collaboration with academic organizations and other thought leaders such as BCnet, Texas A&M, CANARIE, and Internet2. www.apparentnetworks.com
Loki will be speaking at these upcoming conferences: