March 1, 1999
Myths and Misconceptions of IP Telephony
By Ofer Gneezy
Internet protocol (IP) telephony continues to heat up the already smoking
telecommunications industry. As VIP Calling Inc., Burlington, Mass., and other
next-generation telcos play David to the established circuit-switched Goliaths, we
alternately are buoyed by some analysts’ predictions of an "imminent packetized
world" and cast down by others who see only Chicken Little squawking about the sky.
But how sharp are those slingshot stones? And what’s actually holding up the sky?
In the midst of any period of hype there comes a time to clear the air, to cast off the
prevailing tales born from an initial absence of real-world experience. In this spirit, I
offer the following "unconventional wisdom" based on my recent adventures
delivering millions of wholesale IP telephony minutes around the world.
Myth: Voice over IP (VoIP) is inexpensive and low-quality.
Reality: VoIP is inexpensive and capable of delivering toll-quality calls.
We don’t believe inexpensive, low-quality voice is a significant market. The real
market is inexpensive, toll-quality voice. IP telephony service providers have
created networks and monitoring tools to deliver the quality customers expect at
packet-based costs. This quality includes one-stage dialing, no echo, low latency, low
post-dial delay and high call-completion ratios.
Myth: Fax is easier than voice.
Reality: Real-time fax technology is still in flux.
Many industry pundits portray fax as the easy way to enter Internet telephony markets.
Unfortunately, real-time fax actually is quite difficult because it lacks the flexibility
and versatility of voice. Two fax machines operate under extremely tight protocols. This
is partially because if I send you a spreadsheet, you don’t want to hear that it is 99
The industry protocol debate continues: Should fax go transmission control protocol
(TCP) or unreliable datagram protocol (UDP)? The jury, for my part, is still out, but the
end game is clear: Gateways must provide universal port functionality. This way, when a
gateway receives a call, it immediately delivers it to the destination regardless of
whether it was a voice or fax call.
Myth: Worldwide point-of-presence (POP) coverage is the key for every service
Reality: Every IP telephony service provider (ITSP) has worldwide presence.
Every ITSP has worldwide coverage today. By effectively carrying some of our traffic
on-network and some off-network, VIP Calling provides its customers the ability to call
everyone in the world. For instance, if an end user in Japan wants to call London, where
VIP does not have a POP, the traffic comes over the Internet to the United States where we
reoriginate it on the public switched telephone network (PSTN). The "United
States-London" aspect of the call costs less than 4 cents a minute, making the entire
call extremely cost-effective.
Myth: Interoperability is the most important issue facing VoIP providers.
Reality: Operability is more important than interoperability.
By all accounts, today’s first-generation IP telephony hardware lacks scale,
flexibility and reliability. Most gateways have one or two T-spans. This pales in
comparison to the demands created by large call volumes generated through wholesale
operations. ITSPs need port numbers in the hundreds, not tens. Additionally, many gateways
are based on industrial-strength personal computers (PCs), resulting in limited
routing-choice flexibility and extensive integration requirements. Hubs, routers, channel
service units (CSUs), data service units (DSUs), DS-3 multiplexers, digital system
cross-connect frame (DSX) panels and gateways must be assembled to deliver the service
that customers expect.
Truly operable gateways would provide more of a "tandem-switch" offering
reliability and functionality with a high degree of integration. Equipment providers need
to undertake the challenge of assembling components into boxes. Once the equipment is
truly "industrial strength," interoperability will provide real value.
Myth: Value-added services are the key to IP telephony.
Reality: Today’s value is in voice and real-time fax.
There are three phases in the deployment of IP telephony: Phase one is to offer the
same services at lower cost; phase two is to add value-added applications; and phase three
is packet-based network domination. Today, despite all of the hype surrounding value-added
services, the service offering is really voice and fax at a lower cost. Outstanding,
life-changing applications will arrive, but they are not available for commercial
deployment right now.
Myth: Enterprise networks will emerge as the dominant VoIP market channel.
Reality: IP telephony is enabling carriers to provide the enterprise with the
cost advantage that previously was only possible through a private network.
Voice private networks allowed corporations to perform rate arbitrage when there was no
alternative technology to avoid regulation. Wholesale IP telephony delivers comparable
cost-savings with the added benefit of outsourcing network management, which is not a core
competency of many corporations. Many companies currently invest significant resources
babysitting their networks. As the demand for new services increases, the technological
knowledge required will only expand. This will drive the cost and complexity of internally
managing a network upward. The simultaneous new opportunity for cost-effective outsourcing
will decrease popularity of enterprise networks.
Myth: The public Internet is too unreliable to deliver high-quality voice.
Reality: Using the public Internet for Internet telephony is not easy, but the
right tools and processes can deliver toll-quality voice over the public Internet.
Initial PC-to-phone VoIP offerings have colored prevailing perceptions of quality.
These models allowed callers to use the Internet indiscriminately with unpredictable
quality. Commercial operations are not held prey to these vices, however; they can
establish high-speed connections and judiciously choose less congested Internet service
providers (ISPs). Carriers can maintain multiple DS-3s into the Internet backbone and
require that Internet connections into each country support low latency.
Latency, however important, has been misperceived by some as the sole key to unlocking
quality. In reality, a provider must deliver quality across multiple parameters–call
completion rate and post-dial delay echo cancellation, in addition to latency. To achieve
this while maintaining an Internet cost structure requires maintaining "two
networks." VIP Calling does this through assured quality routing, a set of processes
and procedures that automatically reroutes calls to the PSTN when parameters on a given
call exceed developed acceptable ranges.
The final stamp of quality, of course, comes from the customer. Customers simply will
not tolerate poor quality; they will hang up. Average call duration, which is more than
five minutes (for mixed voice and fax traffic), provides the most powerful testimony of
quality. Today’s IP telephony customers are receiving the quality level they expect.
IP telephony pulses with excitement. Every week the Davids of the market announce new
sharp stones while the Goliaths trump aggressive migration plans to avoid the falling
(circuit-switching) sky. The long-foreseen opportunity is real and focused players are
delivering on it now. Assuming regulators continue their current favorable strategy, the
clear winner will be the consumer.
Ofer Gneezy is president and CEO of Burlington, Mass.-based VIP Calling Inc., a
wholesale provider of global Internet protocol (IP) telephony services. He can be reached
at +1 781 229 0011.
Greg Boyd, executive director of the Michigan Alliance for Competitive
"We welcome an SBC-Ameritech alliance. Our experience shows that we can trust SBC
Morton Bahr, president of the Communications Workers of America (CWA)
"The merger would immediately concentrate nearly 40 percent of the nation’s
Economist Joseph Gillan, former staff member of the Illinois Commerce
Read more about:Agents
About the Author(s)
You May Also Like