Perhaps the call for unity in the ADSL world is only a sign of its existing divisions.
On Monday, the Universal ADSL Working Group - officially established itself in the name of speeding the deployment of a consumer version of asymmetric digital subscriber line technology (ADSL). Its high-profile membership includes Microsoft, Intel, Rockwell, Lucent, and Compaq from the PC world, and all the regional Bell operating companies, Sprint, and GTE on the telephone side.
An industry standard for ADSL is key if the technology is to one day resemble current modem technology in ubiquity and ease of installation.
The working group "might start to cull out some decisions," said DataQuest telecommunications analyst Brett Azuma - but the next, and not instant, step is "to build and deploy standards."
In fact, the very need for the working group is a sign that today's ADSL technologies are neither universal nor singular. As with 56-Kbps modems, there are multiple, competing technologies.
On the other hand, 56-Kbps modem technology never saw this kind of high-profile, unified push for a single standard so early on. The group opened its doors with the goal of achieving a universal, single standard - so-called Universal ADSL - for the deployment of consumer-oriented ADSL technology - both in consumer modems and phone companies' switching offices.
ADSL uses standard copper phone lines to speed data between phone company offices and ADSL-equipped PCs. The kind of ADSL sought by the UAWG promises rates of 1 to 1.5 Mbps; today's modem speeds are measured in kilobits per second. The ADSL working group's hoped-for standard works without the need for a splitter at the customer's home, which brings down the top data rates (to 1.5 Mbps or less), but makes this "splitterless" technology much cheaper for phone companies to deploy.
A slew of companies who have trumpeted their own proprietary ADSL technologies are named in support of the working group initiative: among them Aware, Alcatel, Lucent, Paradyne, Rockwell, 3Com, and GlobeSpan Semiconductor.
That's why some observers see new cause for hope that ADSL - nearly written off only a few months ago in the face of momentum for the consumer-oriented cable modem - now holds considerable potential for widespread success.
Competing technologies
But when it comes to standards alone, there are issues to resolve.
The various technologies vying for DSL (digital subscriber line) standards are largely designed around one of two primary technologies: DMT (discrete multitone) and CAP (carrier amplitude/phase modulation).
The former is now tagged as the odds-on favorite for consumer ADSL at the International Telecommunications Union (ITU), the international standards body that will determine the ultimate standards for splitterless ADSL.
The momentum shift for DMT at the ITU began when a group of companies closely resembling the current ADSL working group made a move last December favoring a standard based on the existing ANSI (American National Standards Institute) standard, T1.413, established several years ago by the US standards body and based on DMT technology.
The initiative helped bring about a proposal in the ITU called G.dmt and the related G.lite, which is specific to the splitterless flavor of ADSL that the working group is calling for.
"The line code has not been selected [by the ITU]," said Greg Whelan, director of product marketing for Aware, an ADSL modem vendor whose equipment is based on DMT technology. "But the writing on the wall is that it will be DMT - and it will be interoperable with [higher-speed DSL technologies]."
Proponents say DMT has the edge because of it has demonstrated interoperability among the equipment of different manufacturers, and also because it has a longer reach. All DSL technology tends to degrade in speed across longer lengths of copper wires, but DMT less so than CAP.
Whatever the outcome of this DMT versus CAP debate, however, the ADSL working group and the December initiative by its members have put DMT on top. And thanks to that, Ken Krechmer, communications consultant and US delegate to the International Telecommunications Union, says the ITU may achieve "technical approval" of the proposed G.lite and G.dmt standards, which he says would fulfill the requirements outlined by the working group, by October of this year.
Furthermore, though the ANSI standard originally calls for a splitter - hardware installed where phone lines enter a customer's building - Krechmer thinks G.dmt could be drafted to operate both with and without such hardware.
If the ITU standard was interoperable with splitter-based ADSL, such a single, interoperable ITU standard would be especially attractive to phone companies who put a premium on easy upgradability, Krechmer believes.
With a scalable standard, businesses and telecommuters at home could install or eventually transition to the faster splitter-based ADSL technology without changes by the phone company.
All friends for now
CAP-based ADSL technology has had its adherents, not to mention installed technology. 3Com, for example, has been offering CAP-based ADSL modems since the middle of last year; Ameritech's commercial ADSL offering in the Midwest is CAP-based; as is US West's ADSL service, which employs Pardyne technology. But so far there have been no loud arguments from vendors working with CAP technology.
3Com says its modems, based on DSP chips that are altered with software changes, can fairly simply be adapted to a DMT standard through software.
Bell Atlantic, which had been considered a major CAP holdout, has also changed its tune in the wake of the ADSL working group, of which it is another named supporter.
"We're not hung up on it," says spokesperson Lawrence Plumb. "Early on we favored CAP, now we're not so sure."
Although he says Bell Atlantic would prefer the final standard include, not exclude, CAP as well as DMT, "at the end of the day, we're united behind the idea of accelerating the promotion of standards."
GlobeSpan, once a unit of AT&T and yet another ADSL technology vendor supporting the working group, is the actual inventor of CAP-based DSL. But even if that company should decide to change its tune and make a push for CAP, it would not be enough to derail DMT's momentum, says Krechmer. "With Lucent and Rockwell backing DMT, Globespan is not enough to make a difference," he said.
The way companies supporting the working group are acting, they seem intent on changing the image of themselves as squabblers over proprietary technologies responsible for confusing consumers' purchasing decisions.
Nonetheless, whether it's DMT or CAP, the various companies with proprietary ADSL implementations - Lucent, Amati, Aware, Globespan, Paradyne, Rockwell, and others - are expected by some observers to seek a standard based on their own DMT or CAP variation.
That's where room for potential in-fighting remains, says DataQuest analyst Lisa Pelgrim - in fact, even more so than with 56-Kbps technology, she says, where only two technologies had to battle it out.
"Between technology providers, there will be some very big battles," Pelgrim predicts. "We've seen many different announcements from many different players on a splitterless version of ADSL."
Yet Bill Rodey, vice chairman of the ADSL Forum, an industry group of some 200 communications companies seeking to promote what the forum calls the enormous market potential of ADSL, still thinks the potential for such squabbling is low "because all the heavyweights are banding together."
Rodey extends his faith even to the regional phone companies, who are often pointed to as culprits in the slow deployment of technologies like ISDN and ADSL.
"There's a new access charge that is probably substantial revenue improvement for them. So I don't see how the RBOC is motivated to do anything but get on board."
Only the beginning
Yet no matter how soon a standard is achieved, observers like Krechmer say claims of usable consumer products by Christmas 1998 are at least a year too optimistic.
Some, including DataQuest's Pelgrim, think at least five years is a more realistic timeframe for fulfillment of the ADSL working group's goals.
Thus, the emergence of the working group may be the first good news that has so far been scarce for ADSL. Or as Azuma puts it, the establishment of the group "improves the situation."
"But," he says, "the situation before was pretty abysmal."
