Header
Home | Sitemap Set as homepage | Add to favorites
  Search the Site     » Advanced Search
Sections



The PSTN: The First Hierarchy

by

image

The PSTN gradually assumed its present form over a period of more than 100 years, but almost
from the first it began to exhibit its core architecture. This was based on a point-to-multipoint
topology on the local level, with lines extending out from a local exchange, sometimes called
the central office, and proceeding to larger regional aggregation points that themselves were
linked together in a mesh arrangement that spanned the continent and ultimately the world.
For decades all connections, both short and long, took place over copper wire and involved
mechanical switches that moved pieces of copper wire. Furthermore, all communications took
place in the analog domain.
During the 1960s, fully electronic switching came into being while voice signals began to
be digitized. Otherwise, however, the physical plant remained much the same.
During the last two decades, the physical plant of the PSTN has undergone considerable
modification. The tendency today is to terminate copper-pair phone lines at what are
called digital loop carriers (DLCs), which are neighborhood aggregation points, often mounted  outdoors and utilizing hardened enclosures. The analog voice signals carried by the copper
lines will then be digitized and will be conveyed back to a central office over optical fiber. Dialup
Internet connections will undergo a similar transformation inasmuch as they are actually
carried within analog audio signals.
DSL lines will terminate at a DSL access multiplexer (DSLAM) and will generally also run
over fiber from the aggregation point to the central office. A DLC for ordinary phone lines and
a DSLAM can and often do occupy the same enclosure today.
Today the central office itself is a facility that is occupied by a device known as a class 5
switch whose purpose is to manage traffic over circuit voice channels. The switch’s primary
function is to establish a path to the call’s intended destination utilizing available network
resources, but it must also be capable of handling a multitude of special features such as call
waiting, call forwarding, conferencing, and so on.
Other special calling features, such as caller ID and one-number portability, are handled
by special applications servers that are dedicated to those functions coming under the rubric
of Advanced Intelligent Network (AIN). Basic switching functions are compromised when the
switch’s control plane is invoked to perform unrelated tasks; hence, the use of separate platforms
to support special features has nothing directly to do with traffic handling.
Telephone switches, incidentally, are enormously complex devices embodying millions of
lines of code. Interestingly, most of the functionality enabled by the code occurs beneath the
surface. Such complexity is reflected in the price, which runs into the millions of dollars.
The class 5 switch is involved in directing calls within the local exchange itself and outside
the local exchange to other local exchanges, which could be located anywhere in the world.
The immediate path that the signal takes is most likely to be onto what is known as a fiber ring,
however. The class 5 switch itself does not handle call routing to other class 5 switches. That
function is performed by a class 4 or tandem switch, described next.
Fiber rings emerged in the 1980s and followed the SONET standard in the United States
and followed the closely related synchronous digital hierarchy (SDH) standard in most other
places in the world. The rings themselves always take the form of closed loops, but they are not
necessarily rings in the strictest sense and may in fact meander over the landscape to a considerable
degree. The reason a closed-loop architecture is used is to provide an alternative path
back around the loop if a break occurs in the fiber.
Although some fiber rings encompass only a single metropolitan area and primarily distribute
data services to businesses within an urban core, the larger rings extend out over a
greater metropolitan area or even several such areas; the area may enclose thousands of square
miles. These link many local exchanges and ultimately off-load their traffic onto larger
switches known as class 4 switches that handle calls going to other area codes or country codes.
In some cases, long-distance carriers own the larger rings, but in other cases independents
such as American Fiber own them. The rise of independents in this area constituted a major
trend in telecommunications during the late 1980s and early 1990s. Some cable operators also
own fiber rings—generally the smaller, metro variety.
Fiber rings themselves are evolving, and one tendency that is beginning to manifest
itself—though more in the East Asian markets than in the United States—is the adoption of
packet protocols intended to replace SONET. Chief among these is Resilient Packet Ring (RPR),
which combines the fast restoration capabilities of SONET with the spectral efficiency of IP and
Ethernet. Such packet rings are capable of handling all types of traffic, including voice.
Class 4 switches located within the rings send traffic out over long lines, which in the past
were copper or microwave links but today are almost exclusively optical fiber. The long lines in use today consist of hundreds of strands of optical fiber, each capable of carrying more than
100 frequencies of light. Each frequency in turn can convey minimally 10Gbps or as much as
40Gbps. Thus, total capacity is in the terabits. Such lines are generally the province of long distance
or inter-LATA (which stands for Local Access Transport Area) carriers, as they’re known
in the business, which form the second or third rank in the greater telecommunications hierarchy
depending on whether one designates metro or regional rings as a separate stratum.
At the top of the hierarchy are companies such as Tyco Telecommunications and Global
Crossing, which command international and transoceanic fiber. These companies came to
prominence during the late 1980s and through the 1990s. The submarine cables they utilize are
typically of very large capacity, equivalent to that of the largest continental long lines, and are
designed to provide years of maintenance-free operation.
This neat and somewhat idealized hierarchy has been considerably confused by the emergence
of the Internet, however, as I will explain in the next section. But before that, I will cover
the role that this first hierarchy, the PSTN, will play in the business of the broadband wireless
service provider.
If one wants to provide voice telephony services, even IP voice services, then one must in
most cases either link up with a central office class 5 telephone switch via a device known as an
IP voice gateway or purchase a class 5 switch of one’s own. This is because ultimately the call is
going to a telephone number somewhere else, not to an IP address. This means that the broadband
wireless operator will have to form a relationship either with the incumbent telephone
carrier or with a facilities-based competitive local exchange (CLEC) such as a cable operator
that happens to own a class 5 switch.
In fact, a number of long-distance services use IP as a transport, but, ironically, most of
them also utilize gateways to translate the voice traffic back into circuit form. It is possible to
transmit a call end to end to an ordinary telephone deskset entirely over IP networks, but it is
not the norm today. One would have to establish a special relationship with a long-distance IP
voice service to do so.
Incumbent telcos have offered access to the larger PSTN at a price to other local service
providers for a long time; two-way radio dispatch services are a prime example. If the number
of voice customers in one’s network is small, simply leasing a single T1 back to the incumbent’s
central office may suffice.
In the case of voice telephony services, one always must be concerned with being
beholden to one’s competitor, and any arrangement with an incumbent local exchange (ILEC)
or even a CLEC tends to put one in that position. One possible solution, at least in the case
of facilities-based CLECs, is to attempt to make the relationship synergistic. Some cable operators,
for instance, are starting to view wireless broadband as a means of reaching certain
customers for data services who are not passed by cable currently and cannot be costeffectively
served by this means. Independent mobile operators are another possibility since
they invariably own class 5 switches. A relationship in which the broadband wireless operator
provides access to such customers under some revenue-sharing arrangement in return for a
connection to a class 5 switch at a reasonable rate could be advantageous to both parties.
2013 times read

Related news

» COMPETITIVE LOCAL EXCHANGE CARRIERS
by admin posted on Jun 25,2007
» Beyond the Central Office
by admin posted on Dec 03,2006
» The-Public-Internet-The-Second-Hierarchy
by admin posted on Dec 03,2006
» Voice over 802.11
by admin posted on Aug 23,2007
» Softswitches
by admin posted on Dec 03,2006


More Top News
Cisco Wireless Networking
Most Popular
Featured Author