Convergence /

VoIP: Neither panacea nor pariah

By Kevin Tolly
Network World, 02/18/02

Those who might still wonder whether voice over IP is fundamental to the enterprise or just a fad need only reflect on the volume of discussion to recognize its importance. Joining the fray, they might well find themselves bewildered - as extreme positions are commonplace. To many, voice over IP is either a panacea or a pariah. The truth lies somewhere in between.

While vendor product variations are endless, the voice-over-IP market revolves around deconstructing the monolithic, proprietary, limited-function box known as the PBX and rebuilding that functionality as LAN-based components predicated largely on open standards and open platforms.

This is not evolution, but revolution. And, not surprisingly, there is more than a bit of anarchy and confusion while a "new world order" is being sorted out.

The last time the corporations had to deal with change this profound was in the 1980s when coax-, Synchronous Data Link Control- and twinax-based networks of 3270 mainframe and 5250 AS/400 traffic began the migration to a single, LAN-based infrastructure.

And it's not about saving a few cents per minute on your corporate long-distance bill by transporting telephone calls across your data network. Rather, the new voice-over-IP infrastructure will enable the fabled next-generation applications.

Well, you might ask, where are they? It is a nascent environment. Some have been delivered but most are "on their way." This is part of a natural progression.

Think back. When mainframes were first connected to LANs, the level of integration was almost nil. PCs with 3270 emulation masqueraded as dumb terminals, and "smart" applications meant "screen scraping" data to feed to PC-based applications.

Today, though, there is tight integration across all manner of devices attached to the LAN. The infrastructure needs to be in place before software and application developers have any incentive to be inventive. Voice-over-IP application development will no doubt rise steeply as the number of converged networks increases.

Network architects first encountering voice over IP often mistakenly (but not surprisingly) think that it is monolithic and that there is some well-trodden, universally accepted path to reaching voice-over-IP nirvana. That is hardly the case.

One of the tremendous benefits - and a source of confusion - when building the converged network is that there are a number of equally valid places to begin.

The dramatic "throw out the PBX and replace every phone with an IP phone" scenario is not likely the ideal starting point for most customers.

For many companies, installing an open-platform PBX at small remote offices is the optimal starting point. For others, it might simply be linking PBX and Microsoft directories at the back end or hooking together voice mail and e-mail systems to allow for more flexible delivery of messages.

In every case, a clear business reason can be demonstrated. This is not technology for technology's sake.

With the number of vendors pouring research and development dollars into converged infrastructure and using readily available, high-powered, low-cost hardware and software components to deliver heretofore unimaginable voice-over-IP-enabled infrastructure, converged networks are sure to be the platform for significant competitive advantage in the years ahead.

Recognizing that the industry is still early on in a massive "build out" of new capabilities will help you better understand how to implement your company's voice-over-IP strategy and explain its role to upper management.