Ethernet services

Ethernet Benefits From 21CN Rethink

BT Group plc (NYSE: BT; London: BTA)'s decision earlier this year to rethink its £10 billion ($16.9 billion) 21CN next-generation network strategy has strengthened the incumbent's hand in the Ethernet services sector, where it has made significant progress in an increasingly competitive market.

BT initiated a significant review of its 21CN plans about a year ago, resulting in some major changes in its technology choices, its network and service rollout strategy, and its approach to broadband access. (See BT Unveils $3B FTTx Plan, PBT Sidelined at BT, BT Inches Toward Telco 2.0, and Upheaval at BT's 21CN?)

The revamp was very necessary. As one executive involved in the 21CN process told Light Reading earlier this year, "We were all collectively naive about how tough 21CN was going to be. We all had a naive view on how voice services would migrate, including BT. But the process is industrialized properly now, and the vendors are collaborating better -- there was a lot of finger-pointing before."

He continued: "There's been a wholesale rethink about how to execute. Getting the processes right is the focus now."

The result is that BT has all but thrown out its original plan in favor of a new approach.

At the recent Carrier Ethernet World Congress in Berlin, Tim Hubbard, BT's Head of 21CN Technology and Platform Introduction [Ed note: Does he introduce platforms to each other?], noted that the 21CN project had "shifted focus... Now we are deploying services that our customers want more quickly instead of replicating old [PSTN] services on the new platform."

So instead of focusing on voice service replication and switching off its TDM switches, BT is starting with the rollout of new broadband capabilities, with widespread ADSL2+ deployments (40 percent U.K. household coverage by April 2009), and the delivery of new carrier Ethernet capabilities for enterprise and wholesale services. (See BT Launches Etherflow, Entanet Hops on 21CN, BT Intros WBC ADSL2+, and BT Unveils Ethernet Services.)

Hubbard notes that BT had installed new Ethernet gear –- Alcatel-Lucent (NYSE: ALU)'s 7750 Service Router –- in 106 points of presence (POPs) around the U.K. by the end of May this year, and was due to have 600 POPs installed by the end of May 2009. In addition, he says BT is offering Ethernet services in 174 countries worldwide. (See AlcaLu Expands at BT.)

In the domestic enterprise market, where BT is pushing Etherflow, its newly launched 500 Mbit/s MPLS-based private line WAN connectivity offering, the national operator faces stiff competition. Cable and Wireless plc (NYSE: CWP) will be a much stronger rival once it can reap the benefits of THUS plc (London: THUS)'s Ethernet heritage, and plenty of other players are vying to win corporate and government business with new services and support functions. (See THUS Launches Managed Ethernet, THUS Recommends C&W Offer, and C&W Offers £329M for THUS.)

At the Berlin tradeshow, Exponential-e Ltd. attracted unsolicited admiring comments about its service capabilities, while Virgin Media Business Ltd. unveiled its upgraded customer portal that integrates customer service requests to the service provider's workflow management systems, and enables customers to log faults and check on order progress. (See ntl:Telewest Touts Portal, Exponential-e Touts VPLS Success, and Exponential-e Serves Up Ethernet.)

And then there's Colt Technology Services Group Ltd , which is building out a new pan-European Ethernet network. (See Colt Unveils NGN Vendors and COLT Bolts to NSN Ethernet Switch.)

Wireless backhaul gains
BT is already making significant strides with its new Ethernet capabilities in building up its wholesale Ethernet backhaul business for mobile operators, something the U.K.'s wireless operators have been quick to sign up to. (See MBNL Backhauls With BT, BT Uses Tellabs for Ethernet Backhaul, and BT Wins New O2 Deal.)

Such mobile backhaul services were originally going to be based on the PBT technology that was in favor at BT until earlier this year. Now the services are very much MPLS-based. (See BT Sells PBT-Based Backhaul Service.)

Hubbard explains that BT had requirements to deliver the services quickly to the mobile operators, and "the only way to deliver the service on time was to use Ethernet over pseudowire."

With BT now heavily reliant on AlcaLu for its Ethernet capabilities (and other services), the presumption among many is that the British operator will eventually deploy MPLS-based, connection-oriented technology for easy-to-manage, low-cost packet transport. Currently that technology is T-MPLS, but it is being revamped under a new standards process into MPLS-TP (Transport Profile), a technology AlcaLu is already pushing hard. (See BT Outsources Ops to AlcaLu and Transport MPLS Gets a Makeover.)

But Hubbard says BT isn't committed to any new technology, and, when asked specifically about how BT might deploy PBT or MPLS-TP in the future, he said BT is "still evaluating technologies to see how best they might be used to meet customer needs. What we are not doing is building a platform and then trying to figure out how we might be able to use it."

Amen to that.

— Ray Le Maistre, International News Editor, Light Reading

digits 12/5/2012 | 3:29:42 PM
re: Ethernet Benefits From 21CN Rethink That whole debate is still raging.

There is still a lot of work being done on OA&M specifications for carrier Ethrenet services, and the general view from the market is that significant progress has been made in being able to manage services and deliver SLAs, particularly in terms of the capabilities enabled at the demarcation points at the edge of the network, but there's still a way to go, especially if you compare carrier Ethernet against traditional data services.

That comparison has seen the emergence in the packet transport world of PBT and T-MPLS/MPLS-TP -- two approaches to connection-oriented Ethernet, whereby carrier operations teams would be able to predetermine service connections in a familiar way.

All these topics are going to be examined at next week's Ethernet Expo in New York - see


gbmorrison 12/5/2012 | 3:29:42 PM
re: Ethernet Benefits From 21CN Rethink Five years ago I heard a lot of equipment guys argue that Carrier Ethernet was crazy and would never work because you could not manage the traffic well enough, their was no way you could get guaranteed QoS. How did that change, or has it?
ethertype 12/5/2012 | 3:29:40 PM
re: Ethernet Benefits From 21CN Rethink "PBT and T-MPLS/MPLS-TP -- two approaches to connection-oriented Ethernet"

Not quite, Ray. The reason PBT has been so challenged is that it tried to invent a new mechanism for provisioning connections that ONLY applied to Ethernet. Carriers looked around and said, wait a minute, we already have a very mature and flexible way of creating connections in our networks. It's called MPLS. It would be lunacy to create a brand new infrastructure for setting up and managing Ethernet connections. Let's instead work on a few enhancements to MPLS (i.e. MPLS-TP) to handle those use cases where we'd like to have more static provisioning, for both Ethernet and non-Ethernet. That way we build on what we have and what we know instead of recreating it from scratch.
jlm 12/5/2012 | 3:29:36 PM
re: Ethernet Benefits From 21CN Rethink It's not clear to me that there's really a benefit in using the same technology in the access network as in the metro/core. MPLS is well established in the latter, but some people think its legacy design choices make it a poor future choice, and new access deployments are more open to newer technologies. If multicast matters increasingly (for IPTV) then PBB/PBT is preferable, but otherwise, an simplified MPLS-based approach might be preferable. Both can do the required traffic engineering and QoS provision. An advantage of PBB/PBT in my opinion is its openness to upgrading via newer control planes.
davallan 12/5/2012 | 3:29:34 PM
re: Ethernet Benefits From 21CN Rethink Hi Ethertype:

Your comments are a bit disingenuous...

PBT as an infrastructure can carry anything for which an Ethertype exists. A rather long list...It is not somthing that only applies to Ethernet.

MPLS-TP will largely be a new infrastructure IF you buy into the myth that it will be a simplified and cost reduced version of MPLS. AND if it is a simpified and cost reduced version it will be somthing new and not "what you have and what you know". e.g. new procedures, new OAM...

If it isn't then why does TP exist if MPLS itself can do the job....

As T-MPLS did not really have any uptake, the statement "carriers looked around" does not really hold water for me either...Both T-MPLS and MPLS-TP are vendor initiatives...

So was PBT once, but it has customers....;-)

digits 12/5/2012 | 3:29:33 PM
re: Ethernet Benefits From 21CN Rethink ethertype
I wonder which company you work for...

PBT (PBB-TE) and T-MPLS/MPLS-TP *are* two technologies that enable connection-oriented provisioning of Ethernet transport - it's not a case of 'not quite'.

Yes, MPLS does other stuff too, but if you want to do connection-oriented Ethernet, then those are two approaches.

Please do make the case against PBT/PBB-TE -- that's not something many people have problems with -- but if "Carriers looked around and said, wait a minute, we already have a very mature and flexible way of creating connections in our networks. It's called MPLS...."

then PBT/PBB-TE would not be deployed ANYWHERE and MPLS would be EVERYWHERE, and that's simply not the case.

Obviously that doesn't please many of the MPLS folk who have were lobbying against PBT before hardly anyone had heard of it (which is what made it such an interesting development, by the way) but market choices show that MPLS isn't currently the only answer to all carrier problems.

And that's not my opinion -- that's just a fact.
Sign In