The Voice Over Broadband Chorus
Players sing different tunes with fledgling technology
By Karen Brown
from the April 16, 2001 issue of Broadband Week
Move over, data. Voice is making a strong, if not always cohesive entrance on the broadband service stage. While voice may be a latecomer to the data-dominated broadband world, its message about generating added revenue is making music for carriers and service providers--particularly given the pragmatism of the telecommunications New Economy.
Stephen Gleave, vice president of marketing for broadband voice provider Jetstream Communications Inc., says whether it is DSL, cable or wireless, voice over broadband has an appeal--even if many did not figure it out until lately.
"The possibility of how much you can shove down a single connection is a compelling business angle for a carrier," Gleave says. "It's fascinating to me to see that it actually took little bit of time for people to wake up to that ... that just running a data-only service over a broadband connection wasn't for a lot of carriers going to generate enough revenue to sustain their business model."
The platforms
While the term "voice over broadband" might conjure up images of a market singing in unison, that is far from the case. Within the term there are distinct parts sung by telco, cable and fixed wireless providers.
With fixed wireless on the horizon and cable operators just starting to roll out broadband telephony to mostly residential users, the most aggressive voice over broadband players are on the telco side, particularly in offering combined enterprise voice and data service.
Competitive carriers have been leading that charge, targeting customers in small to medium enterprise, multiple dwelling unit and office parks. The incumbent Baby Bells, meanwhile, only now are starting to look at VOB services.
"I think you will see some announcements and market trials from some big players in the first part of this year, and deployments happening in the second half of the year. So there is a lot going on in that arena behind the scenes," says Jennifer Stagnaro, chief marketing officer for telco broadband voice supplier CopperCom Inc.
Indeed, a spokeswoman for SBC says the telco is eyeing voice over DSL. "We are beginning tests by the end of the year and are looking at deploying within the next year," she says.
Stagnero says CopperCom also is seeing a lot of activity among newer carriers who are meshing IP softswitch technology and broadband access. "There's quite a bit of activity in voice over DSL and voice over T-1," she says. "We see wireless actually starting to emerge."
Similarly, rival Jetstream Communications does the bulk of its business in DSL and T-1 but is dabbling in some wireless options. While hybrid fiber-coaxial cable technology could be a future addition to the company's connection lineup, so far it hasn't fit into the overall business market strategy, Gleave says.
"We kind of looked at that and said, 'You know, this probably isn't our key business right now,'" Gleaves says. "We may get there--I know the ILECs look at the cable guys and say 'Wow, if these cable guys start running out bundled services over cable they could really threaten our long-distance business or our local telephony service.' It may well be something Jetstream views as a viable market, but for now we've got our hands full with DSL, T-1, E-1 and wireless worldwide."
Farthest behind in development but not off the radar in broadband voice is fixed wireless. One such company taking early steps on this platform is Spike Broadband Systems, which is developing a last-mile wireless system based on a multichannel multipoint distribution service (MMDS) scheme. Spike's Continuum system operates in the domestic 2.5 GHz and the international 3.5 GHz space, pumping up to 10 megabits per second to customers.
Cristian Parrino, vice president of marketing for Spike Broadband, says the technology so far is gaining more interest in Europe and Latin America, where competition for broadband service is sparser and the focus is on voice and data delivery. In the U.S., Spike is in talks with several major IXC carriers, but deployment is still at least a year away.
"We are very much focused internationally simply because internationally is where broadband is going to take off first," Parrino says. "The U.S. seems to be a couple of steps behind when it comes to wireless."
The devil in the transport
Equally divided within the voice over broadband world is the method of transport, with factions favoring older Asymmetric Transport Mode versus those backing the up-and-coming Internet Protocol. But there are still problems with both methods, prompting some providers to step back to a well-worn Time Division Multiplexing (TDM) strategy.
In this case, the uncertainty does create opportunity. Terayon Communication Systems, long a mainstay in cable gear, is hopping into the telco pool with its all-in-one Mainsail multiple access unit that can flex to fit TDM, ATM and IP systems, offering customer connection slots for Ethernet, copper DSL, T-1 and even fiber OC-3. Bill Waters, senior vice president and general manager of Terayon's telecom carrier solutions unit, says one of the reasons for this flexibility in transport and connections is the lack of unity in the broadband voice market.
"It's just confusing times right now," Waters says. "People are still wrestling with ATM--is IP here? The answer is typically no. So what do I do in the interim to satisfy my customers and give myself a unique differentiation in the market?"
Moreover, while considered more stable and proven than IP, ATM also poses problems for voice traffic, so much so that some carriers are going back to old-fashioned, fixed bandwidth TDM technologies connections in the last mile. Waters, for example, finds a lot of competitive carriers are frustrated with ATM voice.
"I've heard two stories. One is the cost and complexity of ATM from the edge of the network to the customer is a lot more complicated and difficult to manage," he says. "So what I see is really a key move to go back in the last mile, from the CO or the POP out, and just do plain old T-1. Its tried, its true, its PCM voice, and you don't have to mess with it. It's easy."
But at the same time, there are other carriers that are interested only in ATM, particularly because of its dynamic bandwidth properties, Waters notes. "ATM and TDM are going to coexist for a while, and I bet I have carriers that deploy both," he says. "IP, I think is going to be a while."
When does IP arrive?
Many in the industry echo that sentiment. IP has found a home on high-capacity backbone networks and local area networks, but the point in between--the local access loop--hasn't proven hospitable for the technology.
"Even in the early days when we went to do our research three years ago--all of the telephone companies said if you want to do this, it has to be with ATM, because that's the only technology right now that's going to give us the quality of service to deliver the telephone services that we need," Stagnero says.
That gap is made more complicated by the fact any voice system eventually must connect to the still primarily analog public switched telephone network. "The big question is going to be, 'How do you keep the packet connections working with PSTN connections?'" Stagnero says. "Because that's what service providers are more concerned about because most people are talking over analog lines in the PSDN today and that's not going away any time soon."
Jetstream's Gleave sees a similar stalemate. "Where we are not seeing IP ready for prime time yet is across the access local loop," he says. "At the moment these voice over ATM with DSL or voice over ATM with T-1 or wireless solutions always convert the traffic back to a circuit switch at the central office, so what is the value? You could do voice over IP in that space, but in fact that just gives you less efficiency than voice over ATM, and because you are not end-to-end IP, its sort of why bother?"
Gleave adds that with IP at either end "there is going to be a time of course when IP is going to kind of compress in from both sides."
"IP from a pure 'I'm going to bet my business on it,' is definitely not there yet," Waters agrees. "The only way to solve voice over IP today--you can have good-quality voice but you have to have bandwidth that is unlimited. And so if you are using the Internet, you can just forget it. It is not going to happen."
Making the switch
Another sticking point for IP is the daunting task of replacing legacy Class 5 switches with IP softswitches. Major incumbent carriers rely on thousands of these standard circuit switches to do much of the voice traffic heavy lifting, and replacing them will be massively expensive. The RBOCs have weathered the Internet freeze better than their competitors, but they remain sensitive to such upgrade costs.
"It's a challenge we know we want to pursue because convergence at that level is going to give us so much opportunity for more advanced applications and the end of the day with simpler management and provisioning," Gleave says. "But to get there I think people have started to realize that it was a lot harder than folks have had you believe when the NASDAQ was at 5,000."
Then, too, is the question of using relatively unproven softswitch technologies, which must duplicate the high call volumes and 3,000-odd call features the legacy switches support. "That Class 5 has a lot of code that has been proven out over 30 years or longer, and then you go to softswitches that haven't been around very long," Waters says. "So there is a huge risk here. There's a lot of trialing and testing going on, but why spread the promise of IP when it is a little ways out?"
But there are cases where the IP softswitch can make more economic sense. John Reister, vice president of strategy and technology for Copper Mountain, points out RBOCs moving into new territory and competitive carriers will use this new technology rather than investing in a $5 million to $10 million Class 5 switch. Softswitches also allow call routing features not possible with the traditional switched ATM network, he says.
For example, a new Copper Mountain IP product is a media gateway controller sitting between the customer's integrated access device (IAD) and the softswitch. It can choose which route the call follows based on the number dialed. In local calls, that may mean diverting it back and avoiding the switch completely.
"So instead of having one humongous gateway that all of the IADs have to home to, you might have a gateway that is in San Francisco, and one that is 10 miles south in Palo Alto, or you might be in Denver and Englewood and Boulder--you can spread these things around. And now what happens is the call flows by the shortest path to that particular gateway."
The economy play
With all of these divisions, you'd think voice over broadband is in trouble. Not so, say insiders. While strategies and technology may differ, combining voice and data for more efficient telecommunications offerings is finding a market.
"It's all money," CopperCom's Stagnaro says. "If you look at the copper infrastructure, right now, you can only deliver with the regular phone system one service over each copper line to a business or a residence. If you can deliver up to 24 voice lines in addition to high-speed Internet access over that one pair of copper, then there are cost savings to be had."
Incumbent carriers are facing a shortage of copper line, and the return on investment for what they do install is on the order of 15 to 18 years, Stagnaro says. "So they would much rather put in a voice over broadband solution that pays for itself within the first year that allows them to deliver all of the services they need to a business or residence over the existing copper plant," she says.
In fact, VoB is one technology that may benefit from the recent market recession worries, Jetstream's Gleave says. "This is the most simply compelling business economic proposition I've ever seen," he says. "The focus has gone from line counts to EDITBA to average revenue per user. Voice over broadband absolutely increases your average revenue per user. It absolutely also increases your earnings."
While activity so far has focused on small and medium-sized business, Gleave sees the residential market rising also. While there are no major voice over DSL rollouts as yet, some are in the works, powered by DSL lines belonging to the residentially focused ILECs
"I think you are definitely seeing a focus now toward potentially residential customers," Gleave says. "In turn, that mandates certain requirements on the equipment vendors like Jetstream in terms of scalability of your products, the price of the IAD--that can't be too expensive--and the provisioning and network management, because suddenly you are managing potentially millions of customers, not just a few thousand."
Indeed, Gleave and others say the real market ignition for voice over broadband will come when a major ILEC decides to weigh in. "A big carrier will move ... it has to move at some stage," Gleave says. "It's like penguins on the iceberg. One of them is going to jump off and then they all will jump in."
|
