Titchonline.com

TitchOnline.com
600 South Dearborn Street

Suite 1902
Chicago, IL 60605 USA
1-312-922-3772
titch@titchonline.com

Straddling legacy and

next-generation networks

Kestrel Solutions

Mountain View, Calif.

Founded: 1997

Product: Talon MX

Fit: Provides quick capacity boost with multiprotocol support for metro fiber networks

Hatteras Networks

Research Triangle Park, N.C.

Founded: July 2000

Product: Access Class Ethernet

Fit: Plug-and-play solution to tie together fiber and copper loop access equipment

Convedia

Vancouver, B.C.

Founded:1996

Product: CMS-6000 and CMS-1000 media servers

Fit: Supports numerous types of IVR, DTMF, messaging and network announcement applications from a single processing platform

April 23, 2002

In this issue:

A New Foot in the Door

Convedia: The Common Thread

A New Foot in the Door

The story in brief: ILECs may be watching their pennies, but key business customers still want more bandwidth and increased control over network services. Smaller vendors with quick-response solutions articulate strong and immediate value propositions for ILECs.

Like most business users, enterprise customers expect their needs to be met regardless of the economic challenges their service providers face. Right now, these customers are demanding more bandwidth along with greater control and management over their network services. To keep and grow business with such key customers, incumbent local exchange carriers must respond with solutions—yet do so with fiscal discipline.

“A carrier is not going to roll out a service just for the sake of rolling out a new service,” says Kevin Sheehan, vice president of marketing at Hatteras Networks. “There needs to be something in it for the ILEC. It’s got be able to sell a customer a new service, a different service or multiple services in a way it hasn't been able to.”

Hatteras Networks, which this week unveils its first product, a local loop access system that ties together optical Ethernet and copper platforms, is one of several companies that just a few years ago might have targeted competitive local exchange carriers. Now Hatteras, along with others, including Kestrel Solutions and Convedia Corp. (profiled below), are pitching a bolder message to ILECs: their products are in the same class as anything from Lucent Technologies, Nortel Networks or Cisco Systems – but are cheaper, better and offer much more value.

While the equipment each offers might be different, the common themes are low cost, immediate deployment; incremental expansion in tandem with demand; and easy integration with legacy equipment.

Buying only as needed

This approach is necessary because ILECs are approving expenditures only when demand for service warrants them, says David Yoffie, senior vice president and general manager-systems line of business for Kestrel. “Investment based on forecast [for demand] is something people have a lot less tolerance for,” he adds.

Return on investment has to be achieved in six months, says Grant Henderson, co-founder and executive vice president of marketing and strategy at Convedia. And when the equipment is in place, new services have to be configurable right from the start.

Kestrel Solutions manufactures transmission equipment for metropolitan optical networks and is positioning itself against Fujitsu Network Systems. Convedia sells a media gateway that it positions against Nortel and Lucent gear. Hatteras says its systems combine the benefits of fiber and copper loop access equipment available separately from Extreme Networks, Cisco Systems and Alcatel Network Systems.

With so many CLECs in bankruptcy or liquidated, ILECs have more control over the pace of deployment of new technology and services. But competition gave end-users a taste of what some service providers are willing to do to win business, increasing customer expectations.

And while the field is less crowded, competition is by no means gone. CLECs like Covad and Focal Communications appear to be surviving. Wireless service providers are making inroads into the enterprise sector and even some cable TV companies believe their reputations have recovered enough to attempt to court that sector, too. There’s no going back to the way things were.

Up the learning curve

The new breed of vendor is also further up the learning curve than the first companies that attempted to repackage Windows NT-based office network equipment as carrier class switches. New ventures increasingly feature co-founders who have experience working with technology suited to the carrier market. Hatteras’ Sheehan comes from Alcatel. Kestrel’s chairman and CEO is none other than Marty Kaplan, the former chief technology officer at Sprint. Convedia’s management comes from Nortel and Newbridge Networks.

With this experience comes maturity and more respect for the ILEC issues of network overlay and evolution. Vendor business plans are more sensitive to the longer procurement process and the difficulty of integrating new technology into legacy networks.

Still, as small companies, their market approach is different from what ILECs have been used to. For one, they are not attempting to sell turnkey systems. New vendors are more likely to depend on vendor partners who can be brought in when needed, as opposed to attempting to custom design an entire solution by themselves. In the past, larger vendors have been able to fill infrastructure gaps themselves, often through a custom-built solution. The consequences were higher costs and a degree of vendor lock-in. On the other side of the coin, the economy new vendors offer often means a more piece-meal buildout.

For ILECs, the benefit is more choices. The price has been giving up the cradle-to-grave attention vendors like Lucent and Nortel would provide under long-term multimillion-dollar contracts. To get the most from the new environment ILECs will have to get used to doing procurement the way the rest of the business world does. Here are some hints for ILECs on balancing the need to meet spot customer demand for new services against the current cost-consciousness.

Know the specific service goal

This is more than just deciding, “we’ve got to roll out Metro Gigabit Ethernet service by next month.” Be clear on the service priorities behind the procurement plan. What exactly are customers asking for? How well will you, as an ILEC, be able to deliver it? Clearly communicate the precise service goal of the RFP to both your staff and the would-be vendors. The result will be more detailed bids that will be easier to compare and more difficult for vendors to fudge.

Know your embedded costs

Vendors are going to make the weak economy a strong selling point. They will promise all sorts of savings. Will you have to take their word for it or can you quantify those claims for yourself? What does a square foot of CO space cost in Boston vs. Bangor? Can you compute the dollar value of labor saved or productivity gained at your company when analyzing two different product approaches? If one vendor can deliver a $200,000 product tomorrow and another can deliver a product offering the same functionality for $50,000 next month, will waiting that month save $150,000 in costs, or cost $300,000 in revenue opportunities?

Know the specific strengths and weaknesses in your network

In the next-generation network, there are no good and bad platforms, just platforms that integrate well or poorly. A vendor may come knocking with some proven software for DSL deployment out of the CO, but what if most of your copper runs out of DLC cabinets? Two companies may offer quality next-generation Sonet equipment, but which one works best with your current Sonet gear? Gigabit Ethernet is a great technology, but what if there’s no fiber running to the customers you bought it to support?

Be serious about product evaluation

As a buyer, you have a right to query as many vendors as you think necessary. But once things move beyond the RFP stage and you start bringing products into labs, you’re signaling a serious intent to evaluate and purchase. If, after a thorough shakedown, you are satisfied that the product meets or exceeds your requirements, buy it. If it doesn’t, send your regrets. But don’t footdrag, demanding more and more attention and modifications, then after 24 months, confess you never had the budget in the first place. This does vendors or customers little service. Get a reputation for jerking vendor chains and you’ll be last to hear about any innovative network technology, regardless of your budget to buy.

**************

Convedia: The common thread

The story in brief: Media servers, which provide a common point of signaling and processing for different applications, can help carriers introduce new services more quickly while lowering the cost of deployment.

No one would think of buying one PC for word processing, another for spreadsheets, still another for email and a fourth for a business calendar. Yet if phone companies purchased desktop applications the way they have purchased intelligent network applications, they would have five or six PCs per cubicle.

Need to play announcements? That’s one service node. Need to collect and process DTMF tones? That’s another. Need to provide audioconferencing bridges? That’s a third. Further, when each new application may literally require a forklift upgrade, responding to even the most simple customer service request is a slow, expensive process.

However, competitive and cost pressures are forcing a change in service deployment that’s long overdue, as noted in the story above. Carriers must quickly introduce services, sometimes to small pockets of demand, so they must accomplish that as economically as possible. Enter companies like Convedia Corp.

Breaking up service nodes

Convedia, based in Vancouver, B.C., and some of its equally upstart competitors, offer solutions that free service providers from the proprietary approach to IN long practiced by switch vendors like Lucent Technologies and Nortel Networks.

Traditionally, each IN application requires its own service node to run the OS, processing and network mediation. It is exactly like purchasing a different PC for each desktop application, and just as costly, as prices on service nodes could be as high as $1000 per port.

But Convedia plays in the softswitch environment, with its media servers acting as a common processing point for different TDM and IP-based applications. In a nutshell, that means service providers can deploy a range of new and enhanced services easily, quickly, and inexpensively – and reduce overall cost of network operations while they’re at it.

Convedia rolled out its first product, the 18,000-port CMS-6000 in February 2001. This month the company added a smaller version, the CMS-1000, to the line. These media servers take advantage of the flexible nature of softswitches.

Primarily designed to port applications and signaling between legacy TDM networks and new IP networks, softswitches separate the processing, signaling and applications functions that Class 5 switches consolidate and manage centrally.

With call processing and signaling decentralized, it was only a matter of time before service nodes went the same way. Add the IP dimension the softswitches bring and suddenly there’s a mechanism to manage voice services from a next generation network platform. That’s the functionality Convedia capitalizes on in its media servers.

For example, network announcements, voicemail, conferencing bridges, and applications that use interactive voice response often require similar processing methods and database interfaces. All rely to an extent on automated voice responses that reside in a database. All may need to process DTMF tones to interpret user commands.

Convedia’s CMS media servers manage database look-up and signal processing functions for all these applications, plus several others, including test-to-speech, speech recognition and video streaming and conferencing. Applications themselves can be purchased off the shelf and layered on top of the media server. Standardized protocols, typically the media gateway control protocol (MGCP) or the session initiation protocol (SIP), are used for signaling.

Since there’s little additional hardware required, infrastructure costs associated with new service deployment can be reduced by 50 to 70 percent, says Grant Henderson, co-founder and executive vice president of marketing and strategy for Convedia. ROI is about six months, plus the service provider can deploy more robust services using next-generation IP.

Convedia’s application partners include jNETx, Nextone Communications, Pactolus Communications Software and Shanghai Bell. Softswitch partners include Syndeo, Sonus Networks and Unisphere Networks.

High and low end

The CMS-6000 is aimed at ILECs and CLECs. The CMS-1000, with a port density of 300 per rack, is aimed at independents, applications service providers, carriers and enterprise users.

Convedia is privately held and has raised $50 million in three rounds of financing, the latest coming last October. It has 65 employees.

Convedia’s primary competition comes from SnowShore Networks and IP Unity, both of which are rolling out media servers for the high-end market. SnowShore is scheduled to begin shipping this spring. Two weeks ago that company announced software applications partnerships with BayPackets, LongBoard, and Ubiquity.

Convedia, however, is the only company working with software partners to create specific applications. IP Unity reportedly is building its own SIP-based applications with a development toolkit from Indigo.

Long-term, Henderson says Convedia’s strategy of partnering with applications developers, rather than competing with them, maintains better relations with companies that can add value to media servers much faster than will be possible for rivals who attempt applications development themselves.

For the moment, however, the common operations platform the media server provides is the best selling point. Incumbents can answer demands for greater service flexibility, especially in terms of personal information technology, without getting swept up in the endless internal debate about the direction of network evolution. The strategic benefit is that with each new softswitch and media server that goes into the network, the stronger the IP foundation for voice services grows.

It’s an example of the inevitability of open platforms and voice over IP. At one level, when it comes to VoIP, ILECs are still debating “revolution” vs. “evolution.” At another level, pure economics has already forced some decisions. The fundamental platforms for VoIP are in place at most ILECs. The choice is how best to build on them. The noise of the VoIP debate notwithstanding, just by being in the network, media servers from companies like Convedia give ILECs greater freedom to reposition and redefine themselves and create the new, powerful and differentiated personal information technology services they can truly make their own.

-- Steven Titch

***************

This is the last issue for Charter Readers who have not registered for a trial subscription at TitchOnline.com.

Once you register you will be eligible to subscribe for a full year at our special 50% off rate to ensure you don’t miss a single upcoming issue of our newsletter and “The Sixty-Second Reality Check.” But hurry, this special offer expires April 30.

Stories we’re developing now include an examination of how cable companies are gearing up to play in the voice market – and how their upgraded networks will compete with yours for customers…an evaluation of the role of systems integrators like IBM and EDS in the new telecom value chain and whether they’ll be your partners—or predators….an analysis of the business strategies of new wireless players and why you might borrow a trick or two from them.

Don’t delay! Visit www.titchonline.com now to complete your registration and/or subscription and ensure your continued access to TitchOnline.com and its unique coverage of personal information technology management.

We invite your feedback. Please write to us at feedback@TitchOnline.com with questions, comments or if you'd like us to offer a trial subscription to a colleague. Thank you very much!