Internet Telephony: Hype Or Opportunity?
How Service Providers Can Capitalize On A
Rapidly-Emerging Industry
BY MORDY ROTHBERG AND DAVID
GREENBLATT
The big buzz word in both the telecommunications and Internet industries these days is
Internet telephony. Every week, industry reports are issued regarding the billions
of dollars that will be spent on Internet telephony by the year 2001, ranging from $2-8
billion. The questions that remain are: Where do these figures come from, what is all the hype
about, and how can service providers capitalize on this profitable industry? The original
hype dates back to 1990, when Howard Jonas discovered a way to supply the world with a
U.S. dial tone and cheap international long distance rates. His tiny International
Discount Telecommunications pioneered what is now a $1.5 billion dollar a year industry
international "call back" telephone service. This technology allows
anyone around the world to bypass the high costs of placing international calls from
countries outside of the U.S. With call back, companies and individuals save as much as 75
percent on their international calls by redirecting them through the U.S. and suddenly
these PTTs had competition. PTTs have been forced to lower their rates and the consumer
has benefited by choice of providers and lower rates.
With this background, it is easier to better understand the
market projections surrounding IP telephony. While call back did not have the support of
such huge companies as IBM, Lucent, Cisco, Ericsson, and Motorola entering the market, IP
telephony is quickly gaining worldwide support. If call back can evolve into a billion
dollar industry in such a short time, it seems that there is no stopping the IP telephony
market from becoming an $8 billion industry by the year 2001. Now that the hype is clearly
explained and defined, the second most popular concern is quality of service (QoS). When
Internet telephony was first introduced back in 1995, the only conversations over the
Internet were from one PC to another. Both parties had to be equipped with a multimedia PC
with identical telephony software and be online simultaneously. Aside from the practical problem of these calls requiring prearrangement, the sound
quality was choppy, delayed, and poor. For PC origination/termination, the PC client
software is generally free. Many gateway manufacturers imply that the public Internet can
be used for voice conversations. Although this is probably true to some degree, it is not
reasonable to expect reliable, toll-quality conversations to be done over the public
Internet. This is simply because of the unpredictability of the Internet load.
Conversations that must cross the Atlantic Ocean are often jeopardized by limited Internet
bandwidth. In addition, the connection from the ISP to the Internet backbone could have bandwidth
restrictions that would hurt the conversation.
IDT ameliorated the shortcomings of the industry with the rollout of Net2Phone, the
first commercial service bridging PCs with telephones world-wide. Introduced in July 1996, Net2Phone enabled Internet users to place calls to any
telephone in the world at drastically reduced rates. Many quality and convenience issues
were resolved since the recipient of the call didn't need a computer, and was able to talk
on a regular telephone. Over 350,000 users worldwide have routed more than 15 million
minutes over IDT's PC-to-phone service. But that was not all. Soon afterward, many
companies began offering products and services enabling people to place calls over the
Internet without the need for a computer. Companies like VocalTec, Lucent, and Brooktrout
sold gateways that enabled voice conversations to enter and leave the IP network, but
didn't interconnect with the public switched telephone network (PSTN), while service
providers like IDT and Qwest began selling long distance service over the Internet for as
little as a nickel per minute.
The primary function of the gateway is the voice-to-IP (and IP-to-voice) conversion. If
the conversation originates/ ends on a standard voice telephone, then the gateway is a
physical hardware box which is usually running NT Server or some dialect of Unix. If the
conversation originates/ends on a PC, then the gateway is the PC itself which is running
client software to perform the gateway functions the conversation would be
conducted using the microphone and speakers of the user's multimedia PC. When looking at
Internet telephony from the technical perspective, it is essential to look at the
components of the entire system. This is probably the largest error made by those entering
the IP telephony world just like owning a handset is quite different from being
AT&T, having a gateway or a PC client for computer telephony is only a very small
piece of a big puzzle. The key point is that although IP telephony may be different in the
underlying technology, the many fundamentals learned in the last 100 years of telephony
still apply. If one were to invent a workable helium car, it would probably still need to
have the many features and options that every standard vehicle requires.
There are several manufacturers that sell "gateways." At a minimum, these
gateways take in some number of voice lines (usually 24, 48, or 96) and convert the voice
to IP. Some of the manufacturers sell the hardware and software for a more durable,
industrial-strength solution. Others only sell a CD with the software.
The challenge of a good IP conversation is caused by the synchronous nature of
telephone conversations. The speakers obviously want the conversation to be intelligible.
They also want the sound to be free of break-up and delay. The maximum delay tolerance is
probably less than half a second. A poor compression algorithm can cause delay, but the
"wild public Internet" can even cause a greater and tragic flaw.
At Net2Phone, our phone gateways are placed on dedicated fiber links that can be
controlled by QoS guidance. The cost of setting up world-wide fiber is prohibitive; we were aided by an existing IP
network that we had set up domestically. Various service providers have been looking to
offer Internet telephony services for a number of reasons, including:
- Value-added service to offer the current customer base.
- A relatively easy way to enter the already deregulating telecommunications market.
- The ability to emit a hightech image with little investment.
- An easy way to establish customer loyalty by offering a variety of services.
The most logical types of companies to begin offering such services include ISPs, PTTs,
new emerging carriers, utilities, and cable companies. Investment is minimal, especially
considering the fact that by offering Internet telephony services, the company is
ostensibly becoming a telephone company.
Although many companies have claimed that they are becoming telecommunications
companies, it requires millions of dollars of investment and an inordinate amount of time.
Offering telephony service is not as simple as putting one gateway in Egypt and one in
the U.S. and you can now offer telephony service between those two countries. To offer
telephony services, you need a fully integrated IP network, a real-time billing system,
real-time data base and monitoring system, scalable, 24/7 customer and technical support
as well as a full redundant system. For example, an ISP in Brazil has enough problems
dealing with their customers. So is this a realistic opportunity for these types of
companies? Absolutely.
IDT has developed many different programs for partnerships for ISPs without worrying
about the headaches listed above. IDT provides ISPs with solutions so that they can either
resell PC-to-phone service as well as use its existing infrastructure and backbone to
provide end users with a phone-to-phone service. An ISP can become a telco easily with
little start-up costs and an easy way to build up a customer base and retain consumer
loyalty at a low cost. This business model can be customized
for each vendor, depending on their needs. For example, IDT recently partnered with Naray,
a Korea-based mobile phone company. Although they were in the paging and mobile business,
they did not have existing lines by which they could route calls, especially considering
that the Korean telecommunications market was on the cusp of deregulation. So how can a
company quickly begin offering telecommunications services in a deregulating market?
By
reselling an existing Internet telephony company's services, a corporation can easily
enter the telecommunications market without needing existing infrastructure. Through our
partnership with Naray, they became a full-service telco with all the bells and whistles.
Let's get back to the monopoly issue. Back when call back was invented, PTTs were
adamantly opposed, and did all they could to attempt to ban it in their country. When
Internet telephony was first introduced, the PTTs started flailing their arms again. But
less than two years later, more than a dozen PTTs have made significant investments into
Internet telephony.
The biggest announcement was Deutche Telecom's deal with VocalTec. Others such as France Telecom and Telecom Italia have announced that they are testing
various gateways. AT&T has released a service in Japan and has also announced plans on deploying
services in the U.S. The question is why? Why are monopolistic telephone companies
entering a market that will cut into their profits?
Progressive PTTs realize that Internet
telephony is a reality. There is a considerable amount of cost sensitive customers that
they can recapture. There is a considerable amount of customers that don't make telephone calls due to the
high rates that they can attract. More importantly is that PTTs are also restricted by
regulatory and settlement rates. By using Internet telephony, PTTs can lower their rates
to their customers and even have higher margins than what they traditionally have. They
realize that liberalization is here, and if they don't begin offering new services, they
will lose their customers to new and emerging carriers. It is of utmost importance to them
to either partner with already-established Internet telephony companies or conduct the
research and development themselves. But if the PTTs don't want to get swallowed and lose
their margins and customers, they should think twice before ignoring this emerging market.
In order to set up a complete Internet telephony service, a number of issues must be
addressed:
- Security.
- Pricing.
- Routing/Termination.
- Billing.
- Reliability.
- Standardization.
- Scalability.
The ability to authorize each user is essential when discussing secure platforms.
In the voice telephony world this is done by knowing the location of the phone device
(a home or corporate telephone). For traveling users, a User ID/PIN is used. The authorization function is real-time,
and must have a back-end database for verification and a process for handling what it
feels is a fraud attempt. Due to the nature of hackers and the clear intrinsic cash value
of a toll-call, a gatekeeper without the authorization function is a problem waiting to
happen. As with traditional phone networks, rating and pricing can be incredibly complex
logic with considerations such as the time, day of the week, method of termination (can
the call be completed totally IP or does it need to go to the standard PSTN), etc. In a
real-time billing platform, the rating function is used to determine the amount of time
that the user can continue this call.
The routing function answers the question, "how should the call be
terminated." Considerations include that the call may have to be passed off to a
standard voice network for example, if the terminating user has no IP address. Even
when an IP gateway destination exists for terminating the IP-originated call, the routing
function must do calculations such as the load currently on that gateway. This process
must use tables of the network of gateways to identify the routing endpoint.
As the IP call progresses, full tracking must be done for the purpose of billing.
At minimum, this mandates the ability to produce a CDR record (the basic information
such as the caller, the destination of the call, the duration of the call, etc.). The
Net2Phone platform can supply comprehensive CDR records in real-time to any server in the
network. Comprehensive billing allows for a rich variety of billing approaches, multi-level
account handling, good history tracking and more. In the case of Net2Phone, we applied the
carrier experience of our IDT corporation to the billing function to assure that the
functions available in our IP telephony solution would be as rich as those provided for
standard voice telephony. All technology requires reliability, standardization, and scalability. This is even
more so in the telephony world where there is already voice technology that accommodates
well these three needs. All the components of the system be reliable, and must be free of
the hazards of overheating common in equipment that is on constantly. They must have some
level of built-in redundancy. Critical-path components (such as the central rating, routing and billing platform)
must be fully redundant with automatic switch-over capabilities. Standardization is
critical in IP telephony without it, one can converse only with other gateways or
PC clients of the same manufacturer. The industry is awaiting a full version of H.323
which should ultimately enable interoperability of platforms. Net2Phone already has a
H.323 version; upon the final draft of the standard, we will have a market-ready solution
hopefully the first, standards-based, complete IP telephony solution. Scalability
mainly refers to the ability to handle large numbers of calls simultaneously. Any
technician will tell you that elegant technical solutions often fall on their face when
they are scaled up. Let us point out one area of scalability concern. All digital
solutions are based on some processor and other chips what happens when these chips
reach processing capacity? In fact, if only one processor in the sequence reaches
capacity, the entire process is slowed no scalability!
Every day, another company
announces plans to enter the Internet telephony market. Some are start-ups and others are
existing telcos such as IDT, Qwest, and Frontier. In order to succeed both in the market
and on Wall Street, these companies all need an Internet telephony strategy. No one is
saying that Internet telephony will immediately replace traditional circuit switching.
Internet telephony has its advantages including price as well as avoiding regulatory
tariffs. Companies such as Qwest and Level 3 have announced plans for deploying huge IP
networks. If a company is seriously considering getting into this industry, securing a
guaranteed network with full redundancy is key for success in this suddenly competitive
market.
Mordy Rothberg, EVP, Net2Phone, has been with IDT for 3 years, first spearheading
international callback sales, and then with the Net2Phone division. Rothberg has vast
experience in the international telecommunications arena, and has spoken at various
telecom and telephony conferences about the effects of worldwide deregulation on the
telecommunications industry, particularly as it affects Internet telephony. He can be
reached at [email protected] David
Greenblatt, EVP of Technical Operations, has much IP multimedia experience and is leading
the development of several leading IP technologies at IDT. Mr. Greenblatt welcomes your
feedback at [email protected]. |