Analyst: Copper Is BellSouth's Gold

BellSouth Corp. (NYSE: BLS) is set to award contracts for its planned new triple-play access infrastructure before the end of this year, and the RBOC will opt for an IP-over-Ethernet approach, according to Jefferies & Co. analyst George Notter.

Notter says vendors have responded to BellSouth's 260-page request for proposal – dubbed the IPTV RFP and issued in late August – and that suppliers are due to be picked in the next three months, with initial deployments starting in the second half of 2005. Notter's contacts say that by 2010, about 80 percent of the households in the carrier's turf (about 9.5 million lines) will be covered by the new infrastructure, which will carry voice, video, and data services – including broadcast TV, video on demand, VOIP, gaming, and high-speed Internet access – over a single broadband connection.

But before video delivery vendors get too excited by the RFP's name, the analyst notes that BellSouth is due to issue separate RFPs for specific video-centric delivery systems, such as video servers, and that this process covers the underlying infrastructure needed to deal with bandwidth-intensive, real-time services.

The key to BellSouth's approach, and the potential for a relatively swift deployment, is the carrier's copper plant, says Notter. It is generally shorter than other carriers', and much of it is newer and comprises two or more copper pairs. As a result, BellSouth can use a technology known as "bonded" DSL to provide very high data capacity over its existing cables. Bonded ADSL2+ technology, for example, can deliver up to 20 Mbit/s over 9,000 feet, says Notter, compared with 6,000 feet over a single copper pair.

This means BellSouth can deliver the speeds it needs for the new services over much of its installed copper (covering about 70 percent of its residential lines, reckons the analyst), and won't have to lay much new fiber, something SBC Communications Inc. (NYSE: SBC) is planning to do (see SBC RFP Refreshes Remotes). This will save BellSouth money and time.

But BellSouth will have to upgrade its DSL equipment, replacing the gear in its central offices and remote cabinets with IP DSLAMs that can support Gigabit Ethernet and multiple flavors of DSL. Notter estimates the RBOC will spend between $500 million and $700 million on DSL equipment alone.

So who's in line for a slice of the action? Alcatel SA (NYSE: ALA; Paris: CGEP:PA), as the incumbent DSLAM provider, is a firm favorite, and that company launched its latest DSLAM with an eye toward triple-play hopefuls (see Vendors Claim DSLAM Breakthrough). Notter marks the French giant as the frontrunner, but he believes multiple vendors stand to win a slice of the action over time, including Adtran Inc. (Nasdaq: ADTN), Advanced Fibre Communications Inc. (AFC) (Nasdaq: AFCI), Calix Networks Inc., Entrisphere Inc., Lucent Technologies Inc. (NYSE: LU), Ciena Corp. (Nasdaq: CIEN), and Pedestal Networks Inc. The RBOC will also need B-RAS (broadband remote access server) gear to help support the bandwidth demands of video and broadcast TV streams, and Notter sees Cisco Systems Inc. (Nasdaq: CSCO), Juniper Networks Inc. (Nasdaq: JNPR), and Redback Networks Inc. (Nasdaq: RBAK) as being best placed to win out. Cisco and Juniper are likely to pick up any edge router action too, the analyst believes.

Among the vendors likely to deliver the Ethernet switches needed to multiplex and aggregate the multiple connections from the DSLAMs are Cisco, Extreme Networks Inc. (Nasdaq: EXTR), and Foundry Networks Inc. (Nasdaq: FDRY), Notter believes.

There's also the matter of the CWDM equipment BellSouth will likely deploy between its remote DSL locations and the central offices, and while ADVA Optical Networking (Frankfurt: ADV), Ciena Corp. (Nasdaq: CIEN), and Nortel Networks Ltd. (NYSE/Toronto: NT), among others, will be in the hunt, Notter believes Lucent, as an incumbent optical equipment supplier, is "well positioned" to win here and pick up some services work in the bargain.

BellSouth had not responded to requests for additional comment as this article was published.

— Ray Le Maistre, International News Editor, Light Reading

Page 1 / 3   >   >>
anoori 12/5/2012 | 3:01:21 AM
re: Analyst: Copper Is BellSouth's Gold IN the name God
Dear Sir
we will decide to purchase "IEEE 802.3ah EFM based IP/Ethernet DSLAM" Please send me some more information about this type of IP-DSLAM.
Best Regards
sincerely yours
Frank 12/5/2012 | 1:11:33 AM
re: Analyst: Copper Is BellSouth's Gold I wonder if anyone sees any significance here in the change of platform type in the field, being IP- instead of ATM- based? Wasn't there some wording about this in the Boucher bill being circulated on the Hill. The bill would assign significance to the method of delivery, equating an Information Services Provider status if IP over Ethernet were used, as opposed to being a Common Carrier where TDM-based technologies, including GR-303 support of ATM, is used. Fred Goldstein? Do you care to comment on this? Anyone?

Frank Coluccio
[email protected]
Frank 12/5/2012 | 1:11:33 AM
re: Analyst: Copper Is BellSouth's Gold Bell South's latest move surprises (me, at least) with their endorsement of IP and Ethernet in the First Mile, although no mention is made directly to the IEEE's recent ratification of 802.3ah, which supports both copper and fiber Ethernet to the x in the first mile. Their move to Ethernet marks a stark departure from the other RBOC's strict adherence to ATM PONs and other variations of the full service area network (FSAN) model that employ ATM. What is not so surprising, however, is that they are maintaing their focus on copper instead of fiber as the mainstay of their triple play entry.

Frank Coluccio
[email protected]
optical_optimist 12/5/2012 | 1:11:31 AM
re: Analyst: Copper Is BellSouth's Gold Question for the LR masses:

As a general rule, does the "IP DSLAM" category mean that ATM transport is not being used on the copper lines ?? It is not difficult for a DSLAM (or the OLT at the head end of a PON)to convert ATM cells to Ethernet frames and present a GigE interface northnound, which is the reason for my question.

Another question for the masses:
Does the DSL Forum strictly align with ATM-based DSL or is their room in their world for IP-based DSL and IEEE 802.3ah ??

This is an interesting development for BellSouth, as the RBOCs have invested a lot of effort in driving ATM-based PON (G983.1) and DSL standards over the years.
Frank 12/5/2012 | 1:11:31 AM
re: Analyst: Copper Is BellSouth's Gold During the mid-to-late Nineties a number of Ethernet-based DSLAMs were introduced and employed by BLECs/CLECs/DLECs for their lower costs, but were relegated to being called DSL concentrators instead of DSLAMs, per se. I am unaware of any work being done in the DSL Forum that recognizes the Ethernet variant at this time, although I could be wrong.

Today, however, vendors such as Lucent are manufacturing IP DSLAMs designed to deliver broadband via Ethernet in the first mile, as shown in this release of their V-16Gäó IP DSLAM, here:

rjmcmahon 12/5/2012 | 1:11:30 AM
re: Analyst: Copper Is BellSouth's Gold The bill would assign significance to the method of delivery, equating an Information Services Provider status if IP over Ethernet were used, as opposed to being a Common Carrier where TDM-based technologies, including GR-303 support of ATM, is used.

It seems like a flawed approach to me. What does the method of delivery matter? Shouldn't it be who pays for the build out and the usage of ROWs that determines common carrier or not?

Maybe too many special interests are making public policy...
triple_double 12/5/2012 | 1:11:28 AM
re: Analyst: Copper Is BellSouth's Gold A DSLAM which has 802.3ah Copper chipset should be described as "802.3ah (or EFM)-based IP or Ethernet" DSLAM to distinguish from other ATM-based IP or Ethernet DSLAM.
triple_double 12/5/2012 | 1:11:28 AM
re: Analyst: Copper Is BellSouth's Gold ATM is used on all the DSLAMs except the ones use 802.3ah Copper compliant chipset. ATM is in the ADSL standard. Any DSLAM which has L3 functions such as radius, DHCP etc. said to be an IP
DSLAM (e.g. Alcatel 7302). Any DSLAM, which has no or basic L3 functions, said to be an Ethernet DSLAM (of course must have Ethernet uplink(s)).
Some vendors says their DSLAM is an IP DSLAM just because of their gear has Ethernet uplink(s). This is wrong because when you look at the specifications you might not see much of L3 functions on it.
Copperwing12 chipset is the only CO chipset which has an MII interface (of course along with UTOPIA) for an Fast Ethernet PHY chip. It has an Ethernet-ATM bridge inside. Although you can use the Ethernet interface for the uplink, it does not necessarily mean that it is an IP DSLAM because of just the Ethernet uplink. It should have above L2.
For 802.3ah Copper standard, if you look at its layering, you do not see any ATM at all.
In my opinion the problem is about the 802.3ah Copper is that, DSL vendors are very experienced on ATM based ADSL, ADSL2/2+, VDSL technology and the rate/reach constraints of 802.3ah. May be the third one, currently, could be the integration. (Conexant G24 CO chipset is a 24 port chipset by using 16 chips, whereas TI AC7 CO chipset is a 16 port chipset by using just 9 chips).
fgoldstein 12/5/2012 | 1:11:09 AM
re: Analyst: Copper Is BellSouth's Gold Frank, I am seeing some ATM-less IP-based transport in the field, with newer gear supporting it. Things like the Zhone MALC and the Occams can do voice + data via Ethernet feed.

On the other hand, regulatory special treatment for IP is noxious. Information is as information does. SBC is trying to suggest that if the copper loop touches IP, it's exempt from unbundling! Verizon is saying that if the CO switch has IP in it, it's exempt from unbundling. This is silly and dangerous.

The plain fact is that IP is a lousy protocol, and only succeeds because the ARPAnet used it (Uncle paid) and that evolved gradually (compatibility must be maintained!) into the public Internet. If there were any real comp sci going on, or any real network research being done post-1993 (not counting "patch the TCP/IP Holy Writ Stack nine ways to the Next World"), then it would be more obvious. I've been around long enough to remember when X.25 was just being standardized (badly, to be sure). Having straw horses to knock over doesn't make IP so perfect that the regulators need knock everything else over on its behalf.
stephenpcooke 12/5/2012 | 1:11:07 AM
re: Analyst: Copper Is BellSouth's Gold FGoldstein wrote:

The plain fact is that IP is a lousy protocol, and only succeeds because the ARPAnet used it (Uncle paid) and that evolved gradually (compatibility must be maintained!) into the public Internet.

Frank, RJ, Sorry I didn't get back to you on the other thread, Canadian Thanksgiving...Turkey overload. I have to agree with Fred's comment above. ATM can move information around quite happily at 100% bandwidth utilization where most IP networks are max'ed at 30% and get seriously bad at 60% utilization. It seems that we are all so shellshocked that we are willing to jump on any bandwagon that seems to have some market acceptance whether it makes good sense or not.

I find it strange that engineers of excellent quality are willing to open the previously isolated PSTN to the dangers of the Internet. Had you asked Bill Gates, in his technical instead of marketing role, 10 years ago if Windows was safe against hackers, viruses, worms, etc. I am sure he would have responded in the positive. We have seen how well their 4 major OS releases have faired so far. What makes the telecom industry so sure that we can stop all nasties when we have absolutely no experience in this area? Given that people's lives depend on the systems we have designed shouldn't we re-think this attitude? We, as an industry, have been humbled in no uncertain terms. If we are to regain our standing in society should we really be making such incredibly risky decisions so quickly after we have been burnt so bad?

There is plenty of innovation left in the people on these boards and in our industry in general in areas that do not necessitate IP-based protocols. Things like security and traffic priority, which are bolt-on items requiring equipment re-configuration which few do properly in the IP world, are pretty much assumed in the telecom world mainly due to the isolation of the PSTN. Most telecom equipment in service today was designed with this mindset. The amount of already installed equipment cannot be re-configured and secured in anywhere close to the next 10 years. How are the small telcos all over the world going to cope with such changes and what are their customers going to be exposed to thanks to the decisions of the bigger players? When are Norton AV and Ad-Aware for 5ESS/DMS do out? When will they be OSMINE certified?

I just don't get it guys. If someone would be kind enough to point me at a/the business case for telecom carriers to move to IP-based networks I would be most appreciative.


Page 1 / 3   >   >>
Sign In