SAN DIEGO -- OFC/NFOEC 2009 -- Verizon Communications Inc. (NYSE: VZ) is forming a plan for its next-generation long-haul network, asking vendors to develop a new type of equipment analogous to the packet-optical transport systems (P-OTS) it uses in the metro.
Verizon wants to start deploying in 2011 or 2012, but the gear it wants -- combining optical switching, Sonet, and MPLS-TP into a long-haul system -- doesn't exist. So Verizon has been making the rounds of equipment vendors, suggesting the idea to them.
Stuart Elby, Verizon's vice president of network architecture, brought up the topic while speaking on a panel at the The Optical Society (OSA)Executive Forum this morning.
The discussions with equipment vendors are still at a very early stage. "Probably the fourth quarter last year is when we started socializing it," Elby told Light Reading.
Elby indicated he's talked to Fujitsu Network Communications Inc. and Infinera Corp. (Nasdaq: INFN) (singling them out because they happened to be on the panel with him) and several other equipment vendors. The new box has no name yet; Verizon is generically calling it the "long-haul optical transport product."
"We will be building it as a replacement for our existing equipment, but that doesn't say it's a flash cut," Elby says. Rather, the gear will be folded into the network as needed. That's going to define some of the requirements for the box, because it will have to handle the spans of a few thousand kilometers that are already in Verizon's fiber network.
MPLS-TP comes into play because that's how Verizon wants to interconnect its backbone routers. "The analysis there is that the sweet spot... is MPLS-TP, not Ethernet switching," Elby says.
That MPLS-TP requirement is responsible for one of the pieces Verizon wants that's not invented: "There's nobody that's built a terabit MPLS-TP switching chip yet," Elby said.
Verizon's planned long-haul overhaul is a reaction to the bandwidth surge created by FiOS. Verizon found ways of boosting its metro networks to accommodate FiOS, but now the bandwidth push is reaching back into the long haul, where some equipment has been in place for (technology) generations.
Elby expects Verizon's capital spending this year to be flat compared with last year. "But the areas of investment change, and I'll be putting more investment, I think, at least in terms of new components and semiconductors, into the backbone network."
Last year at OFC'08, Bob Metcalfe asked the Telecoms industry to re-invent itself, claiming that both the existing "telecoms" technologies (e.g. SONET) and Ethernet were no longer suited for the forthcoming ultra-broadband 100G and above.
The question is: why is Verizon asking vendors to develop a product that don't even exist in their labs, whilst there is an opportunity to develop a brand new technology that would comply to Metcalfe' s remarks ?-á
If I'm a system vendor, I won't answer Verizon' s call. I won't invest into something which is a dead-end anyway: Long-Haul networking will soon require more than a few twists here and there.
Ever heard of market disruption ? Why the heck no one in the Telecoms industry ever thought about the iPhone before Apple ? See what I mean ?
I really wish Google comes out soon with its own long-haul networking gear, just to shake you guys a little bit, just as Apple did with the Nokias and the RIMs...-á
And don't tell me about CAPEX, OPEX and all that stuff: Time has come for a true Revolution. Enough status-quo.-á
re: Verizon Rethinks Long Haul Many years of technology evolution experience has shown Tier 1 operators that few-átechnoogies ever actually come out of the network as new ones are rolled in. Its an unfortunate reality that tech apostles need to udnerstand better.-á Embedded services on old technology drive the revenue profiles that feed CAPEX and R&I dollars to fund new tech.-á Box consolidation in the long haul saves real dollars on operations cost simplifying the network and speeding time to market.-á That makes this type of approach viable with or without the unproven new technologies,
Having not seen an RFI for what ever they are talking about, can anyone provide more color?-á Is it Inter-Metro? 100's of Kilometers... ?-á I belive that problem is solved by several vendors.-á Is it really Ultra Long Haul?...as in 1000s of kilometers?
-áThe big challenge there would be the cost of the platform.-á How many is Vz, or anyone likely to buy?
VZ has publicly stated that they are looking for a POTP/POTS that is the long-haul version of the metro RFP from a few years back (Tellabs and Fujitsu won that one). In general, when VZ says "long-haul", they mean something on the order of 1500-2000km.
There are several carriers talking about this type of product, enough so that you can now purchase reports on the "POTP/POTS market" from a variety of analyst sources. The bigger issue will be seeing if the product definitions from any of those carriers correlate at all with other carriers.
At NF/OFC, there were at least 3 presentations from VZ on this topic that I saw. Requirements mentioned in the talks included 40G and 100G line rates, MPLS-TP and OTN switching, 88 wavelengths, and 8 degree ROADMs. For any more details, you'll have to talk to VZ yourself.
----------------------------- MPLS-TP comes into play because that's how Verizon wants to interconnect its backbone routers. "The analysis there is that the sweet spot... is MPLS-TP, not Ethernet switching," Elby says.
That MPLS-TP requirement is responsible for one of the pieces Verizon wants that's not invented: "There's nobody that's built a terabit MPLS-TP switching chip yet," Elby said. ------------------------------
The developments in the Optical Transport Network (OTN) are providing new tools to efficiently interconnect packet nodes (e.g. routers) located at the edge of the core/backbone domain. As soon as most backbone routers or other metro-core packet nodes are interconnected with each other by Gbit/s LSPs, it is worthwile to consider using the emerging Packet-ODUflex connections (FlexWires) instead and have those switched via subwavelength switching. A FlexWire has a selectable bandwidth of n x 1.244 Gbit/s, with n between 1 and 80.
Terabit subwavelength switches are available today and their capacity may double every year in the coming few years. The next releases will support the emerging FlexWires.
So why use a more complex, non-existing Layer2 (MPLS-TP) terabit switch, if you can use the next release of existing Layer1 (OTN) subwavelength switches.
To save this item to your list of favorite Light Reading content so you can find it later in your Profile page, click the "Save It" button next to the item.
If you found this interesting or useful, please use the links to the services below to share it with other readers. You will need a free account with each service to share an item via that service.
(8) | 
Digg
Del.icio.us
Reddit
Tweet This