Jun 11, 2001 12:00 PM,

Fuzzy reception

by Jim Barthold

It's tough to call cable telephony an oxymoron when more than 1 million people use it.

IP vs. switches

An ingredient in CableLabs' DOCSIS 1.1 effort to develop interoperable data delivery specifications, PacketCable will facilitate packetized voice-over-IP next-generation services and is now a part of the worldwide IPCablecom telephony effort. PacketCable proponents hope to begin widescale deployments late next year. Of course, those same proponents had hoped to begin deployments late this year.

“From a technology standpoint, the transport [mechanism], DOCSIS 1.1 and PacketCable are all relatively far along and pretty much ready for deployment now,” says David Horschak, senior product manager of IP telephony for Motorola's broadband communications sector.

Ideally, voice packets would already be sliding down glass and coaxial chutes; however, PacketCable is still under construction. While it's being finished, some cable operators are doing telephony the old way.

“There are several reasons why we didn't wait,” explains Chris Bowick, senior vice president of technology development for Cox Communications, which now has more than 300,000 telephone customers. “There wasn't any other technology available other than what you call — not my words, your words — 100-year-old technology. We looked at it as an opportunity to get into the business and to provide it as a service to our subscribers.”

Cox is usually strategically joined at the hip with Comcast, but the two split over telephony. Comcast is PacketCable's guiding light, although it is finally starting to get some help from other industry players such as AT&T, which is encouraging to CableLabs and to the industry in general, says Glenn Russell, CableLabs' PacketCable director.

Russell says PacketCable's infrastructure is virtually complete, but an economy that has pistol-whipped key vendors has dampened the technical community's enthusiasm to finish things off. “We're still seeing good participation from the people in terms of bringing product in to test,” Russell says.

It's all in how you look at it

Even if PacketCable were to suddenly emerge as a whole beast, there would still be serious philosophical differences among the cable players about how to offer telephony. Those differences color the way they look at deploying the service.

Charter Communications wants to leverage networks that, unlike existing telephone networks, are relatively new and already carry high-speed data services that run on DOCSIS platforms. The multiple systems operator (MSO) is signing up its first customers for VoIP trials in Wisconsin and St. Louis, “not to take the ILECs head on, but to enhance our product offering for our customers,” says John Pietri, Charter's senior vice president of engineering, who also boasts of not owning one switch.

On the other hand, Greg Braden, AT&T Broadband's executive vice president of broadband services and chief technology officer, has more than enough switches. He's now using them to feed about 700,000 telephony subscribers and is adding more telephone subs every day, despite rumors that AT&T might be abandoning its voice efforts.

“There's been a lot of speculation that we'd pull back on that [when parent AT&T breaks apart], which we have no intention of doing,” he says.

AT&T will eventually offer IP telephony but only when it can do so as reliably as with today's switched infrastructure. It's being careful not to strand its switch investment while IP develops. “From what we're seeing in the IP arena today, they have a long ways to go,” Braden says.

AT&T is one of a group of MSOs that will eventually migrate services from switches to IP rather than move straight into the IP arena. Cox is another. In this scheme, a cable operator enters the business using traditional switched telephony, then migrates part or all of its plant to IP and softswitches. In theory, the operator also manages to retain its switch investment while slowly moving the subscriber base to IP.

“Packet-based telephony, and especially pure softswitch-based telephony, is still maybe a year away,” says Mitch Auster, director of product marketing for ADC's cable systems division and a proponent of the migratory method. “[Operators] all know they want to move all the services they want onto IP.”

It's just that in the time in between there are revenues to be gained and subscribers to be signed up using more conventional methods.

The IP conundrum

No matter what version of telephony a cable operator offers, there's yet another decision to be made: Will it be lifeline voice or secondary data/teen line service? Some operators such as Time Warner Cable are uninterested in becoming a telephone company.

“We would not encourage customers to give up their primary line since that's not the business we would propose to be in,” says Ann Burr, Time Warner Cable's executive vice president.

There are a few very good reasons for that. Primary lines need always-on powering. Cable is loath to power its customer premises equipment. And primary line providers are telcos, which means submitting to regulations. Companies such as Time Warner don't want to put out the money to power their gear — when the power goes out, the cable modem and set-top go off, just like the computer and the television to which they're attached — and they certainly don't want to be regulated.

They “want to pass under the regulatory bar,” says Adrian Jones, vice president of technical sales for Terayon Communication Systems. “That subtle distinction is very important because it enables voice to be carried as an unregulated service as opposed to a wholly regulated competitive local exchange carrier (CLEC).”

Therefore, logic would indicate that switch supporters would become CLECs, and IP players would offer second-tier voice services. But logic and the cable industry have never been bedfellows. Both Charter and Comcast hanker to use their IP prowess to offer primary line service, which is “very similar to what the ILECs are offering,” and will “wait for VoIP to hit prime time,” says Charter's Pietri.

Comcast has always had on the high beams when it comes to packetized telephony.

“We've worked very hard to do it,” says Mark Coblitz, Comcast's senior vice president of strategic planning. Comcast has made that work a closely guarded secret after a very public VoIP trial in Union, N.J., with Lucent Technologies. “The Union thing is over. We're doing other stuff that's not announced and that's all I can say,” he notes.

To be fair, the Union trial was probably over the day it started. Lucent was using a first-generation PathStar technology that didn't fit with Comcast's PacketCable direction. That doesn't mean time or money was wasted, though.

“There are two things we learned,” Coblitz says. “Packet telephony can work, and PacketCable is important. When you're looking at trying to do telephone calls, all those pieces can be done.”

More than a voice

PacketCable, for the record, is more than packetized voice. It's a package of specifications that includes billing interfaces and Class features that must interoperate before CableLabs will apply its stamp of approval. Thus, the technology may be vendor-ready, but PacketCable is still a good year away from fruition.

As Comcast's Coblitz says, “there are lots of things we might like to do with voice across our cable networks in which PacketCable becomes the underlying platform, but that doesn't necessarily mean that they're telephone calls.”

Everyone wants a piece of the huge telephony pie that is made all the tastier by peripheral toppings — caller ID, call waiting and call forwarding — that cost so little to provide and reap so much money.

By using cable's networks and the increasing power of computers, PacketCable helps cable operators pocket these revenues in time and at a cost.

“The question is the degree to which a cable operator fully embraces the vision of PacketCable,” says Kinetic's Harris. “If you're just trying to offer voice you're just offering apples-to-apples service to the LECs but at a cheaper price point.”

Offering cheaper telephone service — although hardly just POTS — satisfies some cable operators. After all, says Cox's Bowick, telephony is “a sticky service when combined with video and data,” which means it's a moneymaking proposition.

“Folks like AT&T and Cox are much more pragmatic,” Harris says. “Time Warner and some of the other ones are certainly interested in what PacketCable can do long-term. I think they're just trying to figure out what that is.”

It's not primary line, says Time Warner's Burr. That belongs to the incumbent telephone company. “There's already a second [coaxial] line into the house today. We have no design in being in competition with the local exchange providers for the primary line services,” she emphasizes.

That sort of industry dichotomy, coupled with PacketCable's slow development, raises some concerns. After all, the ILECs — having disposed of the building LECs, data CLECs and to some extent, the CLECs — can focus their laser beams on those pesky cable guys trying to steal away their customers.

AT&T's Braden says he's not losing any sleep over it. “We're not changing our game plan at all, and we would expect that if the RBOCs believe they have more breathing room to focus against us, that's what they'll do. But we'll be ready for it.”

In the end, reliability rules

Telephony, everyone agrees, can be a successful part of a bundled package. Cable's HFC networks, in fact, encourage bundling because they more reasonably accept services than bandwidth-impaired twisted pairs that the telephone companies have used for more than 100 years.

Telephone networks were set up to deliver voice, which they do very well. Broadband cable, because it is so rich, inherently accepts a bundle of different services.

Unless cable follows the old tried-and-true switched services path, however, it runs the risk of offering a poor substitute for what's already there. As an industry that has been hammered for its poor service in the past — even with innovations such as digital TV and high-speed data — cable cannot deliver anything less than what's available from the guys who have been around longer.

“As a lifeline service we need to be as strong as, if not stronger than, the incumbent LECs,” Bowick says (Figure 1).

PacketCable in the future promises that it will match that reliability. Just as DOCSIS drove the birth and proliferation of cable modems by taking data delivery from the pockets of the dial-up providers, PacketCable could someday take telephony out of the hands of the old line telephone operators.

A mantra permeates every CableLabs session: reliability, reliability, reliability. Cable will deploy no technology before it's reliable, which is why it seems the industry isn't moving.

“Even the advocates, like Comcast, admit that [PacketCable] is still several years away,” Harris says. “The other operators [such as] AT&T and Cox that said they want top-line voice revenue now are doing it. It's a smart way to do it.” Figure 2 shows a network setup designed to serve business and residential customers with voice and data services.

Smart today vs. smart tomorrow

Ultimately, those with the patience and confidence in the opportunity to offer voice will wait for PacketCable. In the minds of Comcast and Charter executives, the rewards will still be there when packetized voice services can be reliably delivered over HFC networks.

Others, seeking the quick hit riches of voice telephony, which already have the infrastructure or are unafraid to pay for it, are willing to take the more conventional route and offer traditional voice services today. However, even they will have to someday move to IP, which has some vendors burned by the CLECs, salivating at a new business prospect.

“CLECs were funded by debt. When they spent that money and couldn't get any more, they went away,” says Gary Tauss, founder, president and CEO of Tollbridge Technologies. “Now we're seeing this second wave and we're focusing on the MSOs.”

It's a smart money move for Tollbridge because, as Tauss admits, cable operators are “cash-flow financed” and “financially interested in the business case.”

Whatever that case may be.

“It's a matter of getting to the best plan,” says CableLabs' Russell. “Everybody knows it'll work.”

Someday.