The real telecom revolution? It's all about WANs, not VoIP: less expensive networking will transform the enterprise

America's Network, Nov 1, 2004 by Dan Moffat

While every politician and pundit proclaims VoIP the Obi-Wan Kenobie that gives hope for competitive telecommunications, a much more far-reaching and important revolution is taking place. Wide Area Networking (WAN) is undergoing fundamental changes that will restructure all enterprise telecommunications for the next 10 years.

VoIP is a delivery technology, not a new service. VoIP dramatically reduces carrier equipment costs, but the really exciting aspect is that it is one element of the trilogy of services in the inevitable IP convergence of voice, data, and video.

WANs began in the 70's and 80's with expensive private lines. In the 90's, cheaper computing power became more prevalent, and less expensive frame relay networks were used to tie it all together. Over the next 10 years, IP VPN will become the dominant form of networking.

Cheap microprocessors are now embedded in everything from gas pumps to oil fryers. Fortunately, inexpensive network IP routers linked by a blend of less expensive broadband access choices can tie together all this distributed business intelligence. Distributed computing and less expensive networking mean that we can begin seeing the promised land of converged IP services for voice, data, and video traffic.

VPNs transporting voice, data, and video are the imminent future of enterprise wide area networking. Yet, there are two major obstacles that prevent attaining the "nirvana" of converged "triple-play" IP services.

TRICKY QUALITY

First, quality of service (QoS) is difficult to maintain within and between IP networks and next to impossible over the public Internet. Anytime you transport traffic between IP networks or over the Net, time-sensitive VoIP is limited to "best efforts" QoS suitable only for consumers. Business class VoIP requires sub-200 milliseconds latency and strict jitter parameters. Despite the promises of new IP technologies such as IP version 6.0 and MPLS, QoS control between IP networks is not there yet.

Even more important (and what everyone seems to conveniently forget in their rush to VoIP) is that all converged IP services by definition require broadband access for the last mile. In today's jungle of multiple access technologies, scores of carriers, and thousands of access services and offerings and equipment solutions, building a regional or national enterprise VPN is not for the faint-of-heart. There are few regulatory incentives for telecom carriers to work with one another on standards, shared QoS, or joint offerings.

Most of today's $20 billion of enterprise WAN runs at speeds of 56 kbps costing around $200 to $250 per location. VPNs provide an existing frame relay customer triple the speed at half the cost. This price is a compelling price-performance leap that is driving mass migration to IP VPNs.

VPNs can also provide dial-up customers with multiple locations an "always on" network for roughly the same price they were are paying for their 56 kbps dial-up.

Building business applications that combine voice, data, and video over an existing enterprise VPN is straightforward. VoIP will come in two forms: "best effort" riding consumer broadband and "business class" riding an enterprise VPN.

VoIP! Scmoip! The real telecom revolution is in broadband access and IP VPN!

Dan Moffat is president and CEO of Vancouver, Washington-based New Edge Networks.