If the telecommunication experts are unanimous on anything, it is the conviction
that all forms of communication are gravitating toward a single unified IP network.
There is less unanimity about when this transformation will be complete,
but few expect it to happen overnight. Both customers and carriers have too much
invested in the PSTN to abandon it, and while the industry has solutions for most
of the technical obstacles, the technologies are immature and changing steadily.
Furthermore, the economics of convergence are insufficient to justify the capital
expenditures of anything but an evolutionary transition.
Historically, enhanced features have started in private networks and gravitated
to the public. While the world lingered over ISDN standards, for example, PBX
manufacturers brought its features to the office. The telephones and protocols are
proprietary, but the features are proven and available in every product line. The
same is happening with the transition to IP. International Data Corporation (IDC)
projects that about 1.4 million IP PBXs will be in service by 2008, but circuit-switched
PBX lines will still outnumber IP lines by a factor of 3 to 1. Convergence faces fewer
obstacles in the private arena because enterprise systems have shorter service lives
than public and the customer has better control of the infrastructure.
Convergence in the public network, where switches typically have service
lives of 20 years or more, is a different matter. The PSTN core is moving toward
IP, but class 5 switches remain circuit-switched. ILECs are beginning to offer VoIP
service over softswitches, but the motivation is more to meet competition than
because of any technical advantage of IP. For what it does, the PSTN is good
enough. It meets the needs of most residence and small business subscribers, and
for voice alone it is superior to IP. Several factors are standing in the way of moving
public voice and video to IP. The most important of these is the lack of a suitable
infrastructure. The IP equivalent of the PSTN is the Internet, but it is not suitable
for commercial-grade telephone service. The Internet is designed to be cheap
and ubiquitous, but it is a chaotic model, not because of the behavior of the
network, but because of the nature of the service. You can have low cost or QoS,
but you cannot have both.
Some observers suggest that the QoS issue is overblown, pointing to cell
phones as evidence that people are willing to forego service quality in exchange
for convenience. This observation is accurate as far as it goes, but it ignores two
issues. One is that business-class telephone service demands consistent quality
that the Internet is incapable of providing. VoIP over the Internet can supplement
the PSTN, but it is not a suitable business model to replace it. The second issue is
the fact that every telephone session has two parties. Some users may decide to
forego quality in favor of low-cost Internet connections, which may be suitable for
family and friends, but enterprises cannot afford to impose poor quality on their
customers. The telecommunications industry has elevated transmission quality to
its present state through a long series of technological advances. To diminish it for
the sake of expediency would be a serious mistake.
That said, the industry is evolving toward the converged network. It will be
a gradual transition; one that will occur only after some significant barriers have
been overcome. The obstacles will be resolved in time with more new protocols
and a lot of industry work, particularly on the infrastructure. The Internet could
be adapted for commercial grade telephone service, but several things would
have to change. ISPs, which are competitive and independent, would have to
agree to adhere to service and quality standards. Open interconnection with their
competitors would need to occur and today’s pricing model would have to change.
The Internet is designed and constructed to offer best-effort service at low cost
and that is inconsistent with the needs of time-sensitive applications.
Service quality notwithstanding, a growing amount of voice traffic is moving
to the Internet as service providers such as Skype, Vonage, and AT&T’s CallVantage
offer subscribers service at prices so low that millions are signing up. So what if the
service is not toll quality. If people can call halfway around the world for nothing
over a broadband connection that they have already paid for, some sacrifice in
reliability is a small price to pay. Cable companies with their broadband access
into most residences and many small businesses are in a particularly advantageous
position to offer VoIP service because they have control of the access channel.
VoIP in the public network is a disruptive technology. Although its impact
has been slight so far, it forces the LECs to review their business models carefully.
When a company can come from nowhere and furnish telephone service over a
broadband connection with minimal investment, it causes concern for telephone
companies and regulators alike. Regulators’ traditional modes of taxation and
control are impossible with Internet telephony, at least at this point. In 2004 the
State of Minnesota attempted to impose telephone regulation on Vonage despite
the fact that the PUC’s control does not extend to interstate calls and it is impractical
to segregate VoIP calls. This prompted the FCC to issue an order preempting the
Minnesota order, but the issue has not been put to rest.
In this chapter we look first at the forces that are driving toward a converged
network. We look at services that countries such as Japan, which has a much
higher degree of broadband development than North America, are beginning to
provide and how these motivate the development of the converged network. We
examine the advantages of convergence and the barriers that are impeding its
development. We look briefly at the Infranet, which is an initiative of several
companies to develop an IP-based alternative to the Internet. The Infranet, if it is
successful, may prove to be the enabling factor that removes most of the barriers.
We conclude the chapter with an Applications section that discusses the ways in
which the converged network is available today and considerations in applying it.
No comments:
Post a Comment