February 2000

BY CHRIS DONNER

As the PSTN opens up to new applications and transport mechanisms, changing from the proprietary, closed system of the past, the key idea is to open up the development environment to take advantage of PC form factors and the richness, creativity, and ingenuity of the PC development community to bring about new enhanced services that will revolutionize the ways that we communicate. And while this represents perhaps one of the biggest gold rushes of the coming century, it is important to bear in mind that gold rushes have always broken more people than they have made. The point is not that we shouldn’t pursue the glimmer of the golden promise. For every 100 developers who believe that they are creating the next killer app or enhanced service, it’s likely that only one or two of them will see a measurable return on their investment.

And the thing is, the developers know this. Yet they still want to take a chance. The potential returns are tremendous — we’re talking billion dollar markets here, many of which have remained amazingly untapped in the past 50 years or so. But developers and investors are only likely to take so much risk. The promise of a 100-to-1 or 1,000-to-1 return is appealing, but only if the required investment is within reach — is an acceptable potential loss.

That is why, for all the excitement about IP-enabled networks and the cries for greater bandwidth, no one is really suggesting that we rip out the old network and throw it away to replace it all with new equipment (well, okay, maybe Cisco is suggesting this, but they have a clear reason for being excessive here — it’s called profit motive). No, what most people are suggesting is that the old network be made as “next-gen” as possible gradually — shifting intelligence to the edge and dramatically increasing the speed at the core, making more efficient use of existing transport media (copper, cable, power lines) to overcome “last mile” issues, and so on.

The idea is to make the network as accessible as possible so that no one company or development firm has to bear the burden of building out an entirely new infrastructure. That would be a Herculean task. Instead, ILECs and RBOCs are opening their networks (or being forced to open their networks — this is not always a friendly move) to allow the entry of “prospectors” who are willing to make the necessary investment in their ideas, their particular genius — who are willing to take the risk that the incumbents haven’t been willing to take for many, many years.

UP TIME
But, this all has to be done without shutting down the phone system. As the name of this section suggests, it must be up 24 x 7. Here’s an example: while I have voice mail on my home phone, and while I am willing to accept a twice yearly shutdown of the voice mail system for “upgrades” between midnight and 4:00 a.m. on Sundays, I am not willing to accept such a shutdown of the phone system as a whole. If the incumbents are to have all of these alternative service providers offering enhanced services over their networks — indeed, if the incumbents are to become enhanced service providers themselves — they need systems that can be put in place quickly, brought up and running smoothly and without adversely affecting existing service, and that can be relied upon to provide PSTN-quality service all the time.

TIME-TO-MARKET
So how do PC-familiar developers who are willing to take the gamble, to bet their own and their investors’ money on a dream and a service, get in the door? First, of course, their platforms must be mission-critical, NEBS compliant, five-nines and all of that. But there are going to be hundreds or thousands of developers out there with access to these platforms and brainstorming on the same ideas. So time-to-market becomes crucial as well.

INCREASING THE PACE
And that is exactly what RadiSys is looking to do with their newly announced PACE (PAcket Circuit Engine) application-ready industrial computer platform. RadiSys has designed the PACE platform family to provide OEMs with a common single platform on which to develop multiple applications for the telecom and datacom markets, and they are targeting the enterprise, ISP, and central office markets. OEMs might be able to do this system integration work themselves, but why should they? When they are in a race against time and against their competitors to provide services that are not only always up and running but are also in high demand, why should they waste precious time in designing underlying systems and testing for compatibility between pieces of equipment?

The answer is that they shouldn’t, and RadiSys has heard that answer loud and clear. They are responding with their PACE initiative. The first member of the PACE family to be generally available is the PACE 2000, an integrated subsystem that is ready for use by OEMs of VoIP, SS7 signaling gateways, IVR, and mixed media applications. Included in the PACE system are a CompactPCI chassis, an Intel-based host CPU, various Texas Instruments DSP-powered single board voice gateways, and all the necessary software (OS, RTOS, kernel, algorithms, and API) pre-installed by RadiSys instead of by the OEM or a systems integrator.

Of course, this dramatically decreases time-to-market of the application and allows it to be shipped more quickly than would normally be possible. But it also brings other benefits. For example, RadiSys has designed and tested all of these components to work together in just such a system, reducing the risk of incompatibility problems and system quirks. So the application-ready platform not only decreases time-to-market — it increases the chance that when the product is released that it will be successfully implemented and won’t lead to unforeseen problems in the network. These platforms have been rigorously tested as integrated systems by RadiSys, plus any testing that an individual developer will do on his/her own applications.

So is RadiSys the only company doing this? Well, their approach is certainly unique, but there are other players in this market who are equally concerned with providing mission-critical development platforms that can be deployed quickly and that can be relied upon to perform.

HERE COMES (THE) SUN
For example, Sun Microsystems has their ServiceProvider.com package designed for start-up service providers who are new Sun customers and who are looking for a platform to support their ISP, ASP, or Web hosting business. With this program, service providers are able to acquire the necessary equipment at a significantly reduced price and gain access to key ISVs as well. The FastTrack Starter Solution comes in three configurations, depending on the requirements and business plan of the intended service provider:

1. Entry Level Solution — a cost-effective solution for a start-up that includes four Netra t 1 rackmount servers, Solaris OS, and Solaris firewall software.
2. ISP Starter Solution — an entry-level platform for a start-up ISP targeting 5,000 to 10,000 accounts that includes two Netra t 1 rackmount servers, two Netra 1125 rackmount servers, Solaris OS, and Solaris firewall software.
3. ASP Starter Solution — an entry-level platform for a start-up ASP that includes two E450 servers, three E250 rackmount servers, So-laris OS, and Solaris firewall software.

Like anyone else looking to sell to the service provider market, Sun is concerned with scalability and reliability, as well as time-to-market. And in this sense, one of Sun’s strongpoints is their use of the Solaris operating system. Although Windows is becoming increasingly popular among telecom developers and boasts a tremendous development community in the more traditional IT world, the fact remains that a large percentage of the existing service provider infrastructure — whether you mean the PSTN or the Internet — runs on some flavor of UNIX, and Solaris is one of the more popular flavors.

THE PACK IS BACK
Hewlett-Packard is also a big force in selling equipment and support to service providers, and they have their HP OpenCall Solution Provider Program in place to help ISVs and systems integrators looking to develop and deploy their applications in the public network. This program includes help on the equipment level — such as assistance in designing, financing, and utilizing a system and technical support on everything from HP-UX to Web and NT development.

However, HP’s program goes beyond equipment and technical support to include promotions and marketing assistance, trade show support, a single contact person for all interactions with Hewlett-Packard, and other benefits. This kind of behind-the-scenes support helps ensure that your development keeps moving forward and gets the exposure it needs to be successful.

This combination of integrated products and extended support for developers is one of the benefits that has resulted from convergence and the use of more open development platforms. Ideas are out there, and suddenly it is very possible for developers with a standard IT background to design enhanced services that can actually be deployed in the public network. However, these people don’t have the background in voice and voice networking requirements that they might need to be considered serious players in this space.

The equipment suppliers stand to sell a lot of hardware and software to these developers, but first they need developers who have enough confidence in their ideas and skills to “put their money where their mouth is.” Thus the rise of these kinds of programs and initiatives among traditional equipment vendors.

These are just some of the initiatives I have heard about out there, and I’m sure there are many more that are worthy of consideration. This is a trend that we can expect to see more and more of as partnering and open systems begins to change the old PSTN proprietary model.

Chris Donner is associate editor for Communications Solutions™ magazine. He can be reached for comment at lguevin@tmcnet.com.

On a somewhat smaller scale, but with similar applications in mind, ICS Advent has recently announced the European availability of their Networked Computer Exchange (NCX) platform. The NCX platform is designed to simplify the development and implementation of converged applications using both data and voice networks, but their target market is the enterprise rather than the central office. From the small business to the large-scale enterprise, the NCX platform easily supports applications such as call logging and reporting, IVR, VoIP and fax over IP, Web integration and e-commerce, and anything else that brings together your voice and data networks.

There are currently four systems that comprise the NCX family of products:

• NCX 1000 — an entry-level solution that supports up to 4 telephony cards.
• NCX 2000 — a solution with increased capacity and performance, handling up to 10 telephony cards.
• NCX 3000 — a mid-range platform for business-critical applications that supports up to 15 cards and includes fault-tolerant features.
• NCX 5000 — a platform for the most demanding business-critical applications; ideal for a large installation requiring maximum capacity and fault tolerance.

The NCX series of products uses Dialogic telephony cards, and the entire family of systems comes with Envox’s CT-Studio application generator as well. So not only is the platform ready for use, but you don’t have to worry about hard-coding your applications. CT-Studio uses a graphical environment for CT development, again focusing on decreasing time-to-market while developing tested, mission-critical applications for the corporate enterprise.