BT Guru Picks Holes in NGNs
Addressing an International Telecommunication Union meeting in London today, Reeve, a key figure in the development of BT's 21st Century Network (21CN), praised the ITU for its role in bringing together the work of many different groups around the world and delivering a unified vision of what an NGN should look like and deliver. (See BT Unveils 21CN Suppliers, Bross: More to Come on 21CN, and Wales to Get 21CN First.)
"The ITU has done a great job in finding a global agreement on NGNs. There's a high level of agreement globally about NGN principles" that has helped deliver an "overall architecture for next generation networks and systems, something that has been unheard of before now," says the BT man. He cited the European Telecommunications Standards Institute (ETSI), 3rd Generation Partnership Project (3GPP), TeleManagement Forum, and Alliance for Telecommunications Industry Solutions (ATIS) as organizations that have helped in the ITU's work.
His praise came as Houlin Zhao, director of the ITU's telecom standards body, the International Telecommunication Union, Standardization Sector (ITU-T), revealed the ITU's Release 1 of NGN Standards specifications -- a whopping 900-page document -- at a presentation near London today.
But while the ITU is making headway in delivering specifications that, says Reeve, are practical and applicable to telecom network operators, he says there are still a number of major NGN hurdles that need to be addressed, most specifically in terms of interconnection, wireless, MPLS, and Ethernet.
Interconnection is becoming "a fundamental issue for NGNs," Reeve says. "At the moment there are various interconnect models for different services, but how do you interconnect MPLS-based NGNs? MPLS doesn't support interconnect at the moment -- there isn't enough information in the header, and that's a major challenge."
Reeve says the issue can be addressed by the IPsphere Forum, the industry body (of which BT is a member) formed to continue the work started by 's (Nasdaq: JNPR) Infranet initiative. (See Infranet Becomes IPsphere.) But, "it needs to do things fast, and it needs to feed its work into the ITU."
Next-gen wireless broadband
Reeve says while there are many debates and arguments about wireless technologies, there is common agreement by telecom technologists that OFDMA (orthogonal frequency division multiple access) is the best technology to deploy for broadband wireless in the future. (See All Hail OFDMA!.)
However, he notes, there are a number of different camps arguing about the best route to OFDMA deployment. The 3G mobile infrastructure players see it as an extension of what they have developed, the WiMax brigade believe they're best placed to deliver, and "then there are proprietary solutions such as the Flarion option from Qualcomm. I just wish we could find one way to get to the technology we want," says a frustrated Reeve, who admits he doesn't know which industry body is best placed to deliver a consensus approach.
MPLS and Ethernet frustrations
As he mentioned recently at Light Reading's Future of Telecom Europe 2005 event in London, Reeve has issues with MPLS and Ethernet, believing there is work to be done before either can be regarded as truly carrier class. (See Aggregation Aggravation.)
"It's not possible to trace an MPLS route end-to-end in a standard way. You have to jump through hoops to achieve that sort of visibility, and that doesn't meet our needs for operations and maintenance. It's good to see that the ITU's NGN Working Group 7 is addressing this issue."
Ethernet also still poses too many O&M issues due to the technology's lack of carrier-class management features. This problem is recognized within the Ethernet community, though, which is working on standard specifications. (See Ethernet Growing Pains .)
Reeve also noted the ongoing dilemma carriers face with new fiber deployments -- "bandwidth demands will eventually necessitate fiber-to-the-home or fiber-to-the-cabinet, but it's a difficult economic case" -- and touched on the need for standards focused on IPTV developments. "There are multiple approaches and different architectures, but wherever the IPTV standards work starts, it should end up as an ITU standard."
Reeve and many of his BT colleagues have been heavily involved in ITU standards work, including representation in the Focus Group on NGNs. Now, with Release 1 concluded, the ITU will focus its attention, in its NGN Global Standards Initiative (NGN GSI), on the detailed protocols necessary to deliver multiple services over NGNs.
But where does that leave the U.K. carrier? BT is determined that 21CN, its NGN, will be built using standards-based systems, but the rollout has begun using technology that isn't based on a set of global NGN specifications. (See Wales to Get 21CN First.)
Reeve says BT would "rather have waited for the standards and worked to those," but the carrier has had to forge ahead to meet its business objectives and has had to work around the ongoing standards work. "We've been trying to convince the standards world that the architecture we have for the 21CN is the right one. If you look at what we're doing, there's nothing new in there from a technology viewpoint. We have based 21CN on existing, popular technologies, such as MPLS in the core and IMS. The plan to shut down the PSTN -- that's new."
BT is spending £10 billion (US$17.2 billion) on its new network, which it plans to have up and running nationwide in 2010. (See BT Moves Ahead With Mega Project.)
— Ray Le Maistre, International News Editor, Light Reading