|
JIM BARTHOLD
Vendors, as a rule, do what their customers want. For that reason, you'll hear a lot of vendors these days talking about telephony switches and migratory paths that take HFC net-works and move them gently from more traditional constant bit rate (CBR) telephony methods to more radical Internet Protocol methods.
That's the lip service. Under their breaths, though, these same vendors will be chanting the mantra: "voice-over-Internet Protocol, voice-over-IP, VoIP."
That's where the future - and therefore, the money - is.
It's the primary reason why a traditional hard-wired vendor such as ADC Telecommunications paid $2.25 billion for software-strong startup Broadband Access Systems (BAS).
"We are now in a position to serve a tremendous global growth potential of IP-based service delivery over wired and wireless broadband networks," says Bill Cadogan, ADC's chairman/CEO.
Translated: "BAS had the IP stuff we need to take our networks to the next level, where we need to be to chase the next wave of the future."
ADC is focused in the circuit-switched environment," Cadogan says. "I think BAS brings a strong next-generation data packet capability. That's a point where ADC was very strong on the voice side, not as strong on the data side, but we think we've taken a large step toward remedying that situation with this merger."
It's a merger that is indicative of an industry trend: If you have the capability to move your traditional telephony solutions to IP, move them. If not, acquire that capability, because that's where the market's going, and it's dragging the customers along.
"Generally the marketplace has decided that if they don't own Class 5 switching infrastructure, they aren't really willing to make that sort of investment at this point in time. They just don't believe it has a 15-year life anymore," says Mark Bakies, Cisco Systems' IP communications senior manager.
Of course, there's the flip side. Switches- CBR - are here now. Bakies carefully dismisses this as part of the future equation.
"Maybe they don't think voice-over-IP for residential telephony is here today, but they actually don't think it's very far away, either," he continues.
If the service providers don't think it's very far away, the vendors think it's going to happen next week.
"We believe IP is very competitive with the switch today," says John Boese, Telcordia Technologies VP/GM-next-generation networks. "It's not just IP, it's IP in a next-generation network, where the operational costs are about 50% of what they are in a PSTN (public switched telephone network) environment, and the capital costs are anywhere from 30% to 50% less than a PSTN environment."
Those kinds of cost savings make it attractive for service providers to start thinking about VoIP today, even if it's not quite ready for primetime, Boese says.
"The question is if the network itself is engineered, architected well, along with the protocols that are available in IP," he says. "That's not an impossible task, and that's what the providers are looking at doing. I think you're going to see a lot of activity around the end of the year, not just in cable, but IP in general and voice-over-IP, NGN (next-generation) networks."
That, of course, is the idealistic, everything-works-and-everything's-clean-along the-networks view. More realistically, some service providers have invested heavily in switches and entrenched in the CBR world, and others are realistically hedging their telephony bets on hybrids that bridge the two worlds.
"That varies by the customer," says John Slevin, Lucent Technologies VP/GM-cable access. "There'll be people who will jump right away to an all IP-based solution because they don't want to go out and invest in Class 5 switches in the interim, nor do they want to form a relationship with somebody else to help deliver that function."
Those are the risk takers who will "bite the bullet and go right after an all-IP network," he says.
Telecommunications, though, especially on the cable side of it, is not filled with risk takers. Many of those players have the first dime they made selling HBO through a trapped service.
"You'll get the other people who are going to lower their risks," he says. "Since they have a very solid network, they'll get started with what's there and still re-use those components when they evolve to an all-IP network."
Slevin sees a migratory path that evolves over the next year or two, based primarily on the hybrid approach of using both switches and IP.
"Once they get the hybrid solution working, which gets them to an IP status in the home, they'll start to realize they can roll more of these faster and get the combination of the bundle of IP data and telephony services," he says.
Another thing influencing this timetable is that the incumbent telco providers are not going to sit idly by and watch outsiders come and devour their lunches like unwelcome brigades of ants at a picnic.
"If cable operators are going to wait 24 months, all these alternative competing technologies of the local access or residential market will just leave a very small opportunity for cable operators," says Golan Manor, VP-technical marketing of Terayon Communication Systems' Broadband Voice Systems group.
Manor, too, favors an evolutionary process "where you have voice-over-IP with all the DOCSIS elements on the access part of the network, the cable TV network, and you can leverage the circuit-switched backbone infrastructure for all the other stuff."
That circuit-switched backbone is not only there, it's growing. Not every MSO is committing to IP and the future. Many, most notably Cox Communications and, to an extent, AT&T Broadband, are just as happy to use proven technology.
"Different customers have different needs," says Lucent's Slevin.
Those needs typically include offering all the class services inherent in today's telephone networks in a fastidiously reliable fashion.
"We helped build that infrastructure with things like our (Class) 5Es (switches) that have been running at six 9s reliability," Slevin says. "They're very proven, very solid. Nobody is questioning when they deploy a 5E it's going to be there, and it's going to run very well."
Nobody's also questioning that the same 5E will cost a small fortune and may have a limited technological lifespan, if not product-wise.
"You're buying it from a solid platform perspective," Slevin argues.
That solid platform allows telephony now. The future, though, is in IP, says Chuck Dougherty, Motorola's VP/GM of Voice- over-IP Solutions. It's all how you get there.
"Our focus today, working with AT&T and others, is on the hybrid approach," Dougherty says. "It takes advantage of the Class 5 switch architecture, which is very mature ... and it takes advantage of the DOCSIS access network, which is also relatively mature."
Mixed into that is a local IP access network that has interfaces to the existing Class 5 switches and "looks just like a circuit-switch network," says Dougherty. "Operationally, customers who have a Class 5 switch infrastructure can utilize the provisioning from the Class 5 switch that exists today to provision local services."
While Dougherty calls that solution "elegant and almost transparent to the operator," it's still the half-step between what is now and what will be full IP.
"That's the big service difference between a hybrid approach and a full IP soft switch approach," Dougherty says. "In that approach, the implementation of those class features on a soft switch, which is really a replacement technology for a Class 5 switch. Arguably, that's 18 to 24 months out at a minimum until you're going to have a really extensive suite of class feature son a soft switch full IP implementation."
That timetable might speed up if high-speed data continues to light up operator revenues and juices the construction and maintenance of networks to deliver IP services. That's part of the reason why the industry is pointing toward DOCSIS 1.1 and its PacketCable telephony components.
"It's a bit of a longer-term play, not only for us, but for the whole industry because DOCSIS 1.1 is fairly complex," says Craig Soderquist, president/CEO of cable modem maker Com21.
"For people who want to deliver telephony today, we can give them IP-like services, quality of service, tiered services, cable telephony, with our (proprietary) ATM (asynchronous transfer mode) system," Soderquist says. "For people who want to get into the market now with ATM, we can meet their needs. Long-term, we'll be there with both modems and headend gear for DOCSIS 1.1 and PacketCable."
Again the timetable is "long-term." The vendors are aiming their product releases up to two years into the future. Thus, this year's Western Cable Show will inevitably be populated by numerous demonstrations of IP capability.
"It's a market in transition," says Cisco's Bakies. "People are already moving that way. You have to segment the market to understand what's going on."
Those segments will, in the end, separate those with class 5 switches from those without and determine how quickly you want to jump into a pool that will soon become frenzied with competition.
"Whether you use the initial technologies that are out there today, of which we have the lion's share of the market, or whether you go into IP is really the decision the carrier has to make based on whether they want to obtain market share today or wait until next year, when the IP solutions are a little more viable and available in more volume," says Oscar Rodriguez, VP-marketing of Nortel Networks Cable Media Segment.
"IP will happen anywhere from the beginning of next year into two to three years from now," he says. "Then you'll have a transition in the marketplace that takes telephony technologies over HFC from a constant bit rate or proprietary system into IP."
And, guaranteed, Nortel won't be the only vendor goosing that transition.
The cable industry entered the high-speed data business with a staged approach.
First, it established a DOCSIS standard. Then it certified first-generation DOCSIS modems. Then, having whetted the public taste for broadband data, it started work in earnest on the niggling details of how to reliably deliver that data with different levels of quality and speed. Called DOCSIS 1.1, this is still in the working stages.
The approach worked. Cable jumped out of the high-speed gate and took off. The telephone industry, which had been diddling with its own digital subscriber line (DSL) high-speed standard, was left in the dust and is only now starting to catch up.
Cable does not seem to be following the same course for voice services. Instead of a best-effort model that delivers pretty good telephone service for second and third lines, or even some chancy primary lines, the industry is either concentrating on finalizing DOCSIS 1.1 and PacketCable standards or using more conventional switched technologies to guarantee reliability.
Or at least most of the industry is.
"There are two mind-sets out there," says Chuck Dougherty, Motorola's VP/GM of voice-over-IP solutions. "Some customers are looking at a secondary line offering where it's just an extension to the cable modem offering. There are people who believe there is a business model that supports that."
Others "view it as unacceptable to allow that telephone operator to maintain primary line service and be able to claim their service is superior by doing that," he says. "They feel it's a prerequisite to have your own primary line service that offers as many features and as high a degree of reliability as the incumbent RBOC does to be able to effectively compete."
Jim Chiddix, Time Warner Cable's CTO, falls into the first camp. Chiddix is happy with the reliability of DOCSIS cable modem service and delighted to pull in revenues from voice services using those modems. These phone services, he says, don't have to be primary line offerings, although they can be positioned as pretty close.
"If you're offering second-line service to people who already have cell phones and aren't worried about power outages, which are very infrequent in many markets, then you can have much lower cost basis for very good telephony service." says Chiddix. "We had better places to put our capital than into the relatively slow payback of primary line or lifeline, generator backed-up circuit-switched telephony."
That's why Time Warner is pushing second-line service, fired off DOCSIS cable modems, in Portland, Maine.
"If the power goes off, the cable modem goes off; there's no voice service," Chiddix says. "If you have a cell phone on your dresser, you know you can make emergency phone calls, so what's the big deal?"
"That second line will have to provide the quality you get from the primary line," says Golan Manor, Terayon Communication Systems' VP-technical marketing for broadband voice systems. "In order to generate additional revenues, you have to have the additional features."
That's what PacketCable is all about and why it's so important to wait before deploying early generation or proprietary VoIP gear, says Steve Craddock, Comcast's VP-new media development.
For one thing, early deployments tend to strand capital.
"If you're deploying proprietary stuff, you're not going to be able to standardize your backends," he says. "You can do things, but there are limitations to it."
For another thing, those technology limitations could muddy the marketplace, Craddock adds.
"I'm not going to take a product that maybe I'll go out with as a second line to start with and eventually end up with one of our business units out there selling primary line," he says. "I'm not going to send them out there with rubber bullets. It's just not going to work."
Craddock will wait for quality VoIP standards to be established and is working as a leading member of CableLabs' PacketCable committee work to make those standards happen sooner rather than later.
Chris Bowick, Cox Communication's SVP-technology development, also doesn't care for existing VoIP methods but for another reason. With established circuit-switched networks, he can look down on VoIP as a hobbyist's toy.
"Voice-over-IP certainly is occurring today," says Bowick, "You can slap a microphone on a computer and talk today. It doesn't provide all the class features nor the reliability of circuit-switched telephony today."
Vendors understand that. It's a reason they're scrambling to get into the end-to-end IP space with products that offer interested MSOs an evolutionary gateway through the existing switches.
"Telephony services are going to drive the revenue and earnings for a lot of these carriers," says Oscar Rodriguez, president/COO of Arris Interactive. "As soon as you have the convergence of IP, which will happen anywhere from the beginning of next year into two, three years from now, then you'll have a transition in the marketplace that takes telephony technologies over HFC (hybrid fiber/coax) from a constant bit rate or proprietary solution into IP."
Two or three years? Maybe not for everyone.
"We believe IP is very competitive with the switch today," says John Boese, VP/GM-next generation networks at Telcordia Technologies. "IP in a controlled environment is a very high quality capability for voice, and that's easily demonstrated in the lab. The question is if the network itself is architected well, along with the protocols that are available in IP"
So, while operators upgrade their networks to handle both data and voice, they can wait for full IP capabilities and standards to be worked out, he says.
Telephone companies are like human beings. Big, muscular, professional-wrestling human beings, but human beings, nonetheless. If you prick them, they bleed, and they get angry.
Cable pricked the telcos with high-speed data, slashing into the marketplace with cable modems while digital subscriber line (DSL) standards for twisted pair networks lingered in standards approvals.
After giving cable the early lead, a hurtin' and angry telephone industry is blitzing back with DSL and is, according to some surveys, making progress to regain what it lost.
Now cable has the scalpel raised, ready to cut even deeper into the Regional Bell Operating Company (RBOC) space with voice-over-broadband services.
Delivered over cable's hybrid fiber/coax (HFC) networks, these voice services are already nicking the RBOCs, who are not expected to take this without a struggle.
"Obviously they're going to react, and we know that," says Greg Braden, EVP/COO of telephony operations for AT&T Broadband.
AT&T's goal is to make a quick slash into broadband telephony and get the jump on the RBOCs while they "are in less of a position to react because they still have to prove themselves and earn the right to get into the long distance business," Braden says.
Using conventional circuit-switched technology for its first voice forays, AT&T is working with Internet Protocol (IP) in the local loop and end-to-end strategies as the kinks work out of these in terms of availability and reliability.
Meanwhile, the telcos are tied up in their own twisted pair wires.
"They've claimed for years that their local service is priced below cost, and they lose money on it," Braden says. "It's difficult for them to react to a new residential-based facilities player (with a) lower cost structure and faster operating environment. They can't react on price."
Add to this the traditional regulatory hurdles through which telcos must leap, and Braden sees cable with an even greater advantage than it had with high-speed data.
"From a technology standpoint, what are they going to react with?" he asks. "They don't have an easy technology solution to come in and replace all their circuit-switched stuff with a cheaper telephony capability because their transport medium is still twisted pair."
That transport medium could, and probably will, change, as fiber prices drop and it becomes expedient to move glass closer to end users. Telcos have repeatedly said they will move fiber to the curb and, eventually, to the home. Again, though, cable will be there first, Braden says.
"We already have fiber we're driving to 250-home nodes," he says. "It doesn't take much to go from that point down to the curb. The distance we have to go versus where they have to go to make that happen is a lot shorter."
No matter what they do, the telcos are starting out behind the cable companies because they have been around for more than 100 years and have accumulated regulations and technology barriers that drag behind like Marley's chains.
"They have plant that, via regulation, has extraordinarily long depreciation schedules on it," Braden says. "They either have to convince the regulators to let them raise their depreciation rates, which means the local telephone service for residential customers will go up in price ... or they have to be willing to take humongous write-offs on their books for that depreciated plant."
Nuera Communications is a typical niche company hoping to profit from broadband's move to Internet Protocol (IP) voice transport.
"We're the piece in this broadband network that makes the packet voice operate at very high quality, very high reliability and scale infinitely. We're the voice guys on the networks," says Bill Ingram, Nuera's CEO.
What's atypical about Nuera is that it has some change in its pockets - about $28.5 million - that is fueled in no small part by a $10 million investment by AT&T Strategic Ventures.
That's no small chunk of change, but "$10 million doesn't buy me," Ingram says. "It creates loyalty, but it doesn't create exclusivity."
The money is a security blanket and "a pretty big endorsement that they intend to go with this technology," he adds.
Nuera's product fills a niche bigger telco network builders either can't, won't or don't want to be bothered with. It improves the creation and performance of voice packets that are delivered via broadband pipes to cable modem termination systems (CMTS).
"What we do is simply convert them into digital voice and send them on to a switch, to another gateway, another point of presence somewhere else in the world," he says.
That could be a conventional Class 5 circuit switch or, at some point in the not-too-distant future, a soft switch with access to the wider ranging Internet backbone, Ingram says.
"We make more money," he says. "We sell more stuff when it goes end-to-end (IP). We want it to go end-to-end."
That's not going to happen right away, he says. At least not with the company's primary customer.
"This is pretty radical for AT&T," he explains. "They have customers, they have money, and they have a nationwide circuit-switched network. For them, it's a very appealing step to provide integrated, converged communications to the subscribers' homes and also connect into their switched network."
So, for now, that's why Nuera will convert its packets into formats those switches can understand. For now.
"We're trying to play this galactic shift that's going on," Ingram says. "Everybody's diving almost like a fumble in football. You know only one side is going to get the ball, but it's unclear because the ball's bouncing around."
Right now there are the new broadband providers- Qwest and Williams, for example- and the old ones- AT&T- chasing that slippery pigskin.
"Those are big guys, and they're going to crush us if we're standing around," Ingram says. "But if we're good at something, one thing, the best in the world at one thing, we have a reason to survive and prosper. We're going to be the pure play packet voice guys and supply this to whoever comes up with the fumble."
Back to this issue
|
