BY ANDRE BALE

A few years ago, when Internet telephony was just beginning, the world saw a brand-new paradigm emerge: Data communications technology seemed to be evolving to the point at which we could recreate the telephone network around data-oriented protocols. Why was this significant? There were three main drivers that made the market sit up and take notice: Cost savings, increased competition, and new applications.

SAVING MONEY
In the early days of the market, we were somewhat na�ve about all of these drivers. We thought cost savings would come because “the Internet is free,” and believed that somehow carrying voice over the Internet would make voice free as well. That turned out to be a fallacy: Free networks by definition can’t carry voice effectively. Adding the ability to carry voice (making the network reliable, predictable, and practical) turned out to cost money — and required the building of a new type of data network. Internet telephony is therefore a misnomer: It’s not the Internet that’s significant, so much as the ability to use networks developed for data services to carry low bandwidth voice traffic. Convergence — using protocols, services, and equipment developed for the data market to carry voice calls — is the name of the game today. And convergence works because data oriented networks are growing so rapidly today that there’s capacity to spare for the small amount of traffic that a typical voice call occupies.

The data world recently celebrated two milestones: 30 years ago the first Internet node was born, as two computers exchanged a few bytes of information through a gateway the size of a refrigerator. And now those few bytes of information have grown to a flood that has just outgrown the size of the global telephone network — until today the largest, most complex machine on the planet. That growth continues unabated. And yet voice is still far more important to most businesses than data. The telephone is how we keep in touch and where we spend our money. A typical small business spends approximately ten times as much on voice services as on data services, and surprisingly, the greater the usage of data services in a business, the greater the use of the telephone. We like to talk.

The impact of convergence will come in several ways:

• Combining different traffic types is better than keeping them separate;
• Data communications products and services are evolving far faster than voice; and
• New service providers have business models that are radically more effective than the old telcos.

THE EVOLVING DATA NETWORK
Every business today needs to have a combination of voice and data services. Whether that means using a modem to dial up the Internet and a plain old telephone to make calls, or whether it involves using a T1 line for data and another to connect the phone switch, every business today has to buy separate services. And yet each network is poorly utilized: Much of the time the bandwidth is unused — particularly on the voice side, where an idle voice channel is completely unused. Converged services mean voice calls and high priority data traffic getting guaranteed access to the bandwidth they need, while spare bandwidth is used for background tasks. This is much more effective.

Since the explosive growth of the Internet started, innovations have come thick and fast in the data world, while the voice world has largely been starved of real fundamental developments. IP switching, fiber communications, xDSL, e-commerce, XML, extranets, and VRML have all come from the need to drive the Internet revolution. Service providers looking to build more effective networks for carrying voice are increasingly looking to Internet-derived technologies to provide the cost savings to allow them to compete with established voice providers.

Much of the benefit of convergence comes down to the service provider. Savings for the end user will come from savings passed on from next generation service providers - and these providers are poised to change what it takes to set up your communications services. Unlike the traditional telephone company, which is interested only in supplying local and long-distance voice and data access, newer entrants are focusing on meeting business needs. Such providers are looking to provide a small business with everything needed to set up and compete on a global stage, and are therefore looking to provide the “office in a box.” Next-generation service providers look less like utilities and more like consultants, providing analysis, applications integration, and communications in a one-stop shop. The battle for the hearts and minds of corporate America has begun. AT&T and Sprint  are investing billions in creating a complete portfolio of services for the small and medium business. And new entrants such as 2nd Century Communications, a second-generation competitive local exchange carrier (CLEC) based in Tampa, Fla., are evolving completely new business models to allow them to be the complete provider of solutions for fast growth businesses.

“Fully integrated solutions are the key to meeting the growing needs of the small- and mid-sized business market,” said Charlotte Baker, 2nd Century’s chief marketing officer. “Next generation service providers utilize convergence to deliver more services and more value along with the best price/performance. Once a service provider is able to add the management capability, they will become a true partner to their customers.”

NEW APPLICATIONS
A lesson that the Internet has taught us is that as public communications merge with powerful applications platforms, new business-enabling applications evolve rapidly. Widely available IP-based data communications, combined with simple-to-program HTML-based servers, fueled the growth in e-businesses. In the same way, combining the public phone network with data applications creates the real benefit of convergence — the ability to create powerful voice-enabled applications at a price point that is available to fast growth companies. You can build voice-enabled applications using traditional PBXs and interactive voice response (IVR) platforms, and some large companies have done so. But you could have programmed a mainframe to act as a Web server…

Next-generation service providers terminate their networks not on dumb channel banks, but on highly intelligent edge devices that tie directly into the applications environment on the customer’s premise — and that increasingly allow applications to control the very behavior of the public network. The impact of this will be immense. Let’s look at three example areas — and remember that opening the network frees innovation to create applications that we’ve never dreamed of:

Customer Relationship Management. We use the phone to keep us in touch, and yet it often fails to do so. Call a supplier — and drop into voice-mail hell. Call a coworker for an update, and play phone tag for the rest of the day. Large call centers have deployed technology to direct calls to the most appropriate person, no matter where he or she is located. Call center technology — the ability for an application to keep control of the call path until the caller has reached an appropriate person — has been prohibitively expensive and complex for any small business to consider. Convergence will allow any business to turn the phone system back into a real tool for keeping in touch with customers.

Mobility. In today’s world we’re increasingly flexible about where we want to work. Whether the teleworker is commuting from the bedroom to the home office, or the sales executive is spending the working hours on the customer’s premise, we need to be in touch and a real part of the office environment wherever we are. We want information to follow us rather than us have to chase it. A converged communications system will make sure that the information I need to run my business (phone calls, e-mails, voice mails, faxes, pages, chat, calendar updates) gets to me wherever I am.

Unified Communications. The tools many of us use today for managing communications are unhelpful in the extreme. Using the phone to scroll through multiple voice mails, keeping files of paper fax printouts, separate systems for e-mail, voice mail, pages, and faxes… these are neither efficient nor user-friendly. Unified messaging has been touted for some time as a panacea for all of this, and yet it’s been costly, difficult to install and maintain, and often not very unified. Convergence brings all communications together into one managed system, allowing me to use the phone to pick up e-mails when I’m on the road, to use the PC’s user interface to manage all my communications, and to phone in and schedule a new meeting with my digital personal assistant. Communications should be a tool, and not a daily grind. Convergence helps manage the flood of information in a more effective way.

We used to think that Internet telephony was a great idea. And so it is — but for rather different reasons than we originally thought. On the horizon is the new Internet: A public voice/data communications system that will create powerful opportunities for business-oriented applications. The revolution has only just begun.

Andrew Bale is CEO of Flexion Systems. Flexion is a young company developing a range of access devices designed to allow a small business to simply adopt the kind of high quality communications systems described in this article. Based in San Francisco, CA, and Corsham in the UK, Flexion takes a global perspective on what is needed to assist small businesses to compete increasingly effectively. For more information, visit Flexion’s Web site at www.flexion.com.

BY DR. CHARLES W. K. GRITTON

The quality dimension is attracting considerable attention at almost every level of telecommunications — for wireless operators, it is increasingly being regarded as an important weapon in the battle to reduce churn. In the wireline domain, the theme is one of maintaining voice quality across a new range of disparate services and network technologies — matching the requirement for flexibility and cost against the benchmark of carrier-grade call quality. Voice-over-IP (VoIP) implementations are coming under intense scrutiny, as the industry gears itself up for the transition from legacy to new-generation broadband networks, capable of delivering a wide cross-section of multimedia services in line with consumer demand.

Compared to other network technologies, VoIP applications represent a particularly challenging environment — especially in the context of echo control. With carrier-grade voice quality evidently the target, there are a number of issues which have to be tackled at source, if this goal is to be achieved. Voice quality can vary on VoIP networks tremendously, involving the impact of a number of different factors. The gateway equipment, the phone systems being utilized, the client software, and carrier infrastructure all have an effect on quality. From an echo control perspective, VoIP applications present a particular challenge and have to be tackled at the source, if the increasing buildup of the potential impairments is to be overcome. An IP network requires a substantially higher standard of echo control to be applied than on comparable TDM networks, particularly in instances where the network is operating close to the limit of the available bandwidth.

VoIP IMPAIRMENTS
VoIP applications involve a significantly higher proportion of impairments than TDM — packetization delay, loss, and jitter are compounded by degradations as a result of coding compression (if used). When these effects are combined, the total effect rises significantly to create a considerably more complex scenario. As delay increases, the requirement for echo control gets much higher.

Comparing this to the delays in a typical packet network, the delays can accumulate in a packet transport system, independent of and in addition to transmission delays. The total roundtrip additional delay can easily be 190 ms in excess of the delay experienced with TDM transmission.

The net result is that VoIP applications require a much greater degree of echo control sophistication if toll-grade voice quality is to be maintained. This puts the focus on the role of echo cancellation and the most effective placement of this function in the network. In the context of IP networks involving H.323 terminals, the echo cancellation function is ideally sited in the CPE itself, operating end-to-end — comparable to an ISDN phone application. A prospect for the far future perhaps, but in the shorter term, reality will center around an access or media gateway. This is comparable to using a Class 5 or Class 4 switch — with the echo control function being located in middle of the network, controlling the echo from the end customer.

The prime differences from the echo control perspective lie in the amount of network variation involved. The CPE example is complex in terms of the acoustic echo that the canceller has to contend with, but the switch example is located deeper in the network, and therefore also has to cope with a greater amount of echo path and signal variation. This also introduces the need for more robust algorithms to be deployed, if effective echo control is to be achieved. Achieving and maintaining quality of service (QoS) is going to be fundamental in the successful rollout of VoIP networks, especially if the technology is going to compete effectively in the medium term. Delivering carrier-grade voice quality is therefore a paramount consideration — and one that cannot be overlooked at the early network implementation stage.

One of the fastest deployment avenues for IP networks will be over access and media gateways, involving a backbone transport role. In this context, there is no doubt that the role of echo cancellation will be an important one, combined with the ability to offer a range of ancillary call quality improvement facilities and performance monitoring capabilities. With the spotlight on ways of optimizing network efficiency and performance, there is clear potential for the deployment of techniques such as noise reduction and level control that can be incorporated cost-effectively and space efficiently from the echo canceller platform. Delivering comparable voice quality to the existing wireline services is the biggest challenge carriers face. For VoIP applications, voice quality will, without doubt, be one of the most significant deciding factors.

There is a strong argument, therefore, for scrutinizing the issue of effective echo control very closely indeed — it holds an important key to attaining this objective.

Dr. Charles W. K. Gritton is chief technical officer for the Network Enhancing Technologies Solutions Group at Tellabs, Inc. Tellabs' transport, access, and service enhancement solutions are used by the company's customers as they provide services to their customers. The products satisfy a dual objective, allowing Tellabs' customers to offer new revenue-generating services while reducing operating and capital costs.