What is telecom’s next great thing

An unprecedented downturn has depressed the industry’s fortunes and could change the way it adopts new technology – at least temporarily.

New technologies come and go in the telecommunications market, but rarely in unpredictable fashion. The life cycle of any new technology can be summarized in three words: hype, bust, reality.

By: Carol Wilson
America’s Network

But as the entire industry suffers from the post-hype hangover left by a massive market crash, it seems unlikely that any equipment vendor or network operator will again boldly venture into the land of overstated claims.

Looking ahead

This condition raises two obvious questions. First is whether the industry has gotten over its reliance on over-promotion to implement new technology. Is hype gone for good or is it merely on an extended vacation?

Second is the more serious question: What’s next for telecommunications?

Outwardly, it seems that innovation has gone the way of the successful IPO, when it fact, it may simply be keeping a lower profile. Many within the industry see shifts, such as the move to use multi-protocol label switching (MPLS) to enable IP networks to carry voice and video traffic, as a major trend that continues without the fanfare.

Equipment vendors such as Juniper Networks and Lucent Technologies are still launching major initiatives, but they are careful to couch their new technology presentations in the language of transition and to pay careful attention to the near-term business case.

For example, Lucent recently announced a new version of its 5ESS switch that has three times the capacity in a small footprint – and will have both IP and SIP interfaces to integrate into an IP-based softswitch network.

"Eighteen months ago, I wouldn’t have thought about including a 5E switch in this story," says Roger Heinz, vice president of product management for convergent solutions at Lucent. "We would have said, ‘Scrap what you have and do it all new.’ But with the emphasis on the business case going forward, we can let them do packet trunking on a 5E and make it much more cost-effective."

In December, Juniper put forth a new reference architecture, the Model for Integrated Network Transformation, that would create an independent service delivery layer within the network for data services

The idea, however, is not to convince network operators that one massive change must be made, says Chris Heckart, vice president of marketing at Juniper. "You have to separate the end goal from the technologies you use to get there," she says. "Getting to one network infrastructure without many overlays is a good goal. We can swap out the individual technologies on the way to achieving that goal. So the first thing is to get broad industry agreement on what the goal looks like."

Why hype?

The need for broad industry agreement to drive economies of scale in developing new technology often creates the need for hyperbole in promoting the next big thing.

To overcome the natural lethargy of the very large companies that operate extensive public networks and to rally the industry around a developing standard or technology requires fanfare, often excessive.

"The carriers, as well, once they decide to go into something, need to create enough momentum with their new services to make it seem these services will be inevitable, so enterprise customers will buy into them," says Charles Giancarlo, senior vice president and general manager of switching, voice and carrier systems at Cisco Systems.

That has come to mean glossing over the complexities of adding anything new to the network.

"The fact is that deploying any new technology in the mass market is hard," says Scott Nelson, Alcatel’s chief technology officer for North America. "It’s very easy to bring something in as a niche play pretty quickly, but it’s a lot more difficult to take it to everybody for the mass market. Mostly a mass market play is about high volume and very low cost, and the key issue there is operational cost. It’s a nontrivial task."

Getting to scale requires integration into public network operations systems that are massive and notoriously slow to change. That’s one of the reasons that "new technologies don’t achieve the cost reductions they promise or prove to be as easy as promised," says Greg Wortman, vice president of marketing at Coriolus Networks.

"It may be easy for service providers to lament the fact that there is technology seemingly available that would allow them to streamline their operations, but implementing that technology is like the sheer difficulty of changing the airplane’s engines in mid-flight," says Scott Steele, vice president of product strategy and portfolio management at Tellabs.

Handling the new and improved

Also working against the deployment of new technology is the continued evolution of existing systems.

"Anyone who is evangelizing new technology never fully comprehends how the incumbent technology will evolve to take on the current threat," says Mike Nielsen, chief technology officer of Lucent’s Integrated Network Solutions.

"That’s what happened with next-generation Sonet and SDH. They continued to evolve to beat out alternatives such as resilient packet ring because the installed base of equipment in the network is Sonet, and it’s easier to only pay for what is needed to transport the new service," he says. "If you look beyond the capital cost to the cost of training, field personnel and other operations, it’s an enormous investment to bring in something brand new. The improvement has to be several orders of magnitude to be justified."

When the telecom bubble burst, it destroyed many incentives to move to new technology platforms, such as the softswitches that were supposed to replace digital voice switches by now.

"The economic advantage of going circuit to packet has diminished over the past few years," says Frank Wiener, vice president of business development at Calix. "That’s one of the things that I think softswitch technology, in particular, is struggling with right now."

Jump start

Given those obstacles and the current lack of capital it almost seems logical for network operators simply to stand pat. But most industry observers believe nothing could be farther from the truth.

"If [network operators] continue at a slow pace to deploy new technologies, they will see declining revenue," Heinz says. "There is also steady competition, including a movement forward with IP PBXs, wireless competition and cable telephony. There are many competitive threats."

Juniper’s Heckart: Separate the means from the end.

The challenge is to prepare today for a dramatically different climate that isn’t likely to occur until 2005, says Tom Nolle, veteran industry analyst and president of CIMI Corp. "The next big revenue influxes aren’t going to come around until 2005," he says. "Carriers have to figure out how to put gear in today that is useful in today’s revenue environment and will continue to be useful beyond 2005, when the service base is going to change dramatically."

The focus on cost-cutting has increased interest in technologies that address some of the operational problems, especially the need to collapse networks by eliminating overlay networks with their separate operations systems and the need to cut the costs of speedy service provisioning.

"New revenue streams are absolutely important, but one of the big barriers has been the cost of provisioning new customer services," says Michael Ladam, analyst with Stratecast Partners. "So you need an infrastructure that is going to reduce the cost of provisioning subscribers to the point where potential services suddenly become profitable."

These technologies include multi-protocol label switching (MPLS) and generalized MPLS, which brings the MPLS protocol set to optical networks, as well as automated switched transport networks and automated switched optical networks

"These protocols have the promise of being able to support bandwidth on demand," says Greg Mumford, chief technology officer of Nortel Networks. "That fits the vision for the network to become more self-service."

Nolle cites AT&T’s Project Pluto as the industry’s first RFP that looks for operations economies of scale to include eliminating the network’s most costly element – humans – whenever possible. That is the only hope for making advanced broadband services cost effective enough to deliver down market – to smaller businesses and consumers – and still make a profit, he says.

Service, no smile

There is also increased focus on developing the next set of services – some would say the first set of services – to use a new converged infrastructure.

"With all the talk about collapsing networks and converging architectures, I think we are trying to solve a problem that doesn’t exist yet," says Chris Nicholl, analyst with Current Analysis. "We keep talking about next-generation services, but there is no demand for next-generation services. We still haven’t come up with the application that will drive consumers to broadband Internet access."

Nolle believes that network operators are caught in something of a service vice. Today’s profitable service products – voice and legacy data such as ATM and frame relay – aren’t growing. The one service that is growing – IP data – remains unprofitable.

Developing data services that will be profitable is a priority but must be accompanied by technology initiatives that will let network operators offer these services widely and profitably, says Juniper’s Heckart. That’s not possible today because of the multiple networks and overlays that have been deployed.

http://www.americasnetwork.com/americasnetwork/article/articleDetail.jsp?id=44012