3.4
Categorize WAN technology types and properties
Today's organizations use many different types of WAN
connections. Your decision to use one connection type over another will largely
depend on the properties each connection type offers and the advantages it may
provide your organization. In the following sections, I'll discuss the main WAN
connection types and their properties.
Most WAN connections are provided by a communications company
referred to as a service provider. One of the main
differences between a WAN and a LAN is that you generally don't own all the
connections on a WAN, unless you are the service provider. Because of this, the
types of network connections you will encounter in a WAN environment are very
different from the connections you are used to in a LAN environment. Most
network connection types have evolved over time, offering the right properties
to connect a company's computers based on the needs of the company for its
connections. I'll discuss WAN connection types such as Frame Relay, T1, T3, DSL,
cable modem, ATM, SONET, wireless, and ISDN.
Let's say your company has one location in New York and another
in San Diego. For your communications, you would like to have a cable that
connects the two locations. You could get a very big truck and a whole lot of
cable and just start driving cross-country, carefully spooling out the cable and
telling everyone not to bother it. Of course, it still wouldn't work when you
got to the end because of the attenuation of copper wire, but that's another
story!
This ridiculous example will help you see what service
providers have done for you. In essence, they have already rolled out that
cable, but they require payment to let you connect to both ends of it. Once you
do, that's your connection, and nobody else is on it. The service provider will
generally “condition” the line from time to time, testing it and making sure you
are getting what you are paying for, but you will be the only one authorized to
use it to communicate. Sometimes this division is only logical, but often it is
physical as well; in other words, it's your wire and only your wire at many
points. Cool, huh?
The most common of these types of connections in the United
States has been the T1. In Europe, they use a very similar
connection called an E1. A T1 provides for 1.54Mbps of
dedicated bandwidth for the customer. This bandwidth can be used in total or
divided up into as many as 24 channels, called DS0s, which are each 64Kbps. An
E1 is very similar but offers 32 DS0s instead of 24 for a total of 2.048Mbps.
This is accomplished using a device called a CSU/DSU. This gives the customer
(you) many options in regard to the dispensation of the bandwidth to network
resources. The cost of a T1 or an E1 varies based on the connection points you
choose, but it's safe to say you would pay between $500 and $1,000 a month for
one T1 line in most locations today. This is why an organization will consider
other options before spending the money on a dedicated T1 connection. I should
also mention that a T1 is not (by far) the most expensive connection type you
can choose.
Caution, the next couple of options are big bucks! What if you
wanted a line like a T1 but much, much larger? Some large companies require
high-bandwidth, dedicated connections from one office or data center to another.
One way to accomplish this is by using a T3 line in the
United States or an E3 line in Europe and much of the rest
of the world. This type of line provides a tremendous amount of usable bandwidth
that can be divided to fit an organization's needs. A T3 provides for 672 DS0s,
or the equivalent of 28 T1s or 44.736Mbps! It is sometimes also referred to as a
DS3. An E3 provides for the equivalent of 512 DS0s, or
approximately 17 E1s or 34.368Mbps.
If you thought those were fast, you ain't seen nothin' yet! In
the term OC-x, the OC stands for “optical carrier,” and the x indicates the
relative speed of the link. Well, the x just keeps getting bigger and bigger.
The original speed of an OC trunk was about 50Mbps, and it was called an OC-1.
OC-3 quickly followed with 150Mbps. The standard at the time of this writing is
OC-3072, which offers a mind-blowing data rate of 160Gbps--that's with a G! As
you might imagine, only very large companies have the need (or the money) for
these options.
How about a protocol that will push data at 150Gbps over fiber
links and has to be controlled using atomic clocks? That's what Synchronous
Optical Networking (SONET) can do. It's especially useful for networks that span
multiple geographic regions because the atomic clock mechanism in it keeps
everyone on the same exact millisecond. As you can imagine, this protocol
requires expensive equipment and expertise. It is typically used by large
communication providers and very large corporations as a transfer mechanism or
backbone for data traffic. In regard to speeds, SONET generally provides an
STS-1 link of 50Mbps bandwidth and multiples thereof, whereas STS-3 would
provide about 150Mbps bandwidth.
Synchronous Digital Hierarchy (SDH) is a standard that is very
similar to SONET. It transfers data over optical fiber using laser light or LED.
Its speed and capability are similar to SONET, and its timing is controlled
using atomic clocks. SDH was originally defined by the European
Telecommunications Standards Institute, whereas SONET was originally defined by
the American National Standards Institute. In regard to speeds, SDH defines an
STM-0 at 50Mbps bandwidth.
Dense Wavelength Division Multiplexing (DWDM) is an optical
technology used to increase bandwidth even further over existing fiber-optic
backbones. It works by combining and transmitting multiple signals
simultaneously at different wavelengths on the same fiber. One fiber is
essentially turned into multiple “virtual fibers.” Currently, because of DWDM,
single fibers have been able to transmit data at speeds up to 400Gbps.
Earlier, when I was talking about “coaxial cable that can
connect you to the Internet,” some of you might have been thinking “What cable?
We live out in the countryside and don't have those cables out here!” If that's
the case, then your best (and maybe only) option is a satellite hookup. You may
also decide to use a satellite hookup because it is the most economical or
dependable service in your area. In either case, you will need a dish antenna
and a professional installer or instructions on how to find their satellite with
your dish antenna. It always makes me laugh when I hear someone say that they
put the satellite in the backyard to get the TV signal. That's not the
“satellite.” The satellite is in geostationary orbit high above the earth, and
your little antenna is just going to pick up on its signal from space.
Once you do that, you can then use that signal as a download
from the Internet. Now uploading is a bit trickier, since you probably won't
have a high-powered transponder. Many satellite communications companies have
provided the upload through your regular telephone line dialup connections. Some
now provide a DSL line to give you more bandwidth for uploads. Others advertise
a two-way satellite system that actually does send some signal back to the
satellite from the antenna. These are typically more expensive and harder to
install. Satellite communications companies offer data rates that rival those of
their biggest competitors, cable and DSL.
This is one that I can't believe is still in the exam objectives! It's been so
long since we used ISDN that it should just go away, and we shouldn't need to
know the details anymore. Oh well, since it's listed on the objectives, I'll go
over the most important aspects regarding ISDN BRI and ISDN PRI that you might
need to know for the test…and very unlikely for real
life.
ISDN BRI is a layer 2 protocol that allows for two
communication channels and one control channel. The communication channels are
referred to as B (bearer) channels, and the control channel is referred to as a
D (delta) channel. Each of the B channels can carry up to 64Kbps of data (that
used to be a lot), and the control channel can use 16Kbps for data control. ISDN
BRI is sometimes referred to as 2B + D, but this is actually misleading because
the B channels are really the only usable bandwidth for data. Thus, an ISDN line
can carry a whopping 128Kbps of data! In other words, one T1 is the rough
equivalent of 12 ISDN BRI lines. ISDN also employs all kinds of telephone
company terminology that identifies the reference points and the devices, but
the chance of you having to know that for this exam (or for real life
anymore) is so remote that it hardly bears mentioning.
ISDN PRI came out a little later, and it's a very different
story than ISDN BRI. An ISDN PRI link is almost the same as a T1 line in regard
to its capacity to carry data. It consists of 23 B channels (each with 64Kbps)
and 1 D channel (also with 64Kbps). A little quick math should tell you that a
single ISDN link will carry 1,472Kbps, which, as you can see, is very close to a
T1.
So, why would a company choose an ISDN PRI link over a T1?
Actually, it could come down to the availability in an area, the cost, the type
of equipment that the company already owns, and the business rationale for the
link. Both ISDN PRI and T1 links offer 23 DS0s for a customer's actual bandwidth
use. (That's the “dirty little secret” about T1s--you don't get all 1.544Mbps
for your use, but you should still know that number for the exam.) Some equipment has
ISDN PRI interfaces built in, which allow flexible control of the 23 B channels,
so a company can use them for special needs such as video conferences or
network-based meetings. ISDN PRI also employs a myriad of telephone company
jargon that, thankfully, you will not need to know. Just know that it's 23 B + D
and that all channels are 64Kbps DS0s, and you will be fine.
Cable companies have jumped on the bandwagon and now offer you
a path to the Internet that begins with connecting your computer to a special
cable modem. That modem is configured by the cable company to be recognized by
its central office, also called the headend of the cable
company. From there, the cable company becomes your Internet service provider
(ISP), connecting you to the Internet.
Many small businesses and home users have chosen this option
for their Internet connection. The advantages include the fact that most cable
companies can provide tremendous bandwidth (10Mbps and faster) downstream and
“acceptable” bandwidth (over 1Mbps) upstream. For most users and small
businesses, this is all they really need. A potential disadvantage is that you
will share bandwidth with others who are in your immediate service area. This
means that at peak times your performance could become degraded. This is not the
case with the next technology that I will discuss.
In the early 1990s, when ISDN BRI was just coming out, I was
working at Sprint in Florida. We used to joke that ISDN stood for either “It
Still Does Nothing” or “I Still Don't kNow.” Sprint wanted about 90 bucks a
month for it so that you could combine your Internet service with your telephone
services. The problem was that most people at that time were saying “What's an
Internet?” You see, the Internet at that point was accessible only by text
commands and usually only through a university or a large corporation. It wasn't
all that exciting then, except for a few of us “geeks.”
By the mid-1990s, though, with the ushering in of the World
Wide Web and hypertext browsing, everything had changed. We sold a lot of ISDN
BRI during that time until the newer technologies of ADSL and cable modem
Internet took its place. Then we still sold it as backup lines to companies that
were leasing a T1 or a T3. It was no longer an expensive add-on by that time.
Now ISDN BRI is all but gone, but you still need to know the historical
information covered in this chapter for the exam.
Another inexpensive option for small companies and home users,
which provides considerable bandwidth for an economical rate, is a digital subscriber line (DSL). The most common of these is
the asymmetric digital subscriber line (ADSL). The reason
this type of connection can be economical is that the lines it uses are already
in place. They are your regular telephone lines. The service provider uses the
regular telephone lines and special equipment that multiplexes the signal to
provide tremendous bandwidth over that which dialup lines provide.
Now, as you may know, asymmetric means “not the same on both
sides” or “not balanced.” So, what is not balanced about an ADSL line? The
bandwidth is not balanced; in fact, it's not even close. You may have noticed
that service providers advertise ADSL using megabits per second for download speed but
kilobits per second for
upload speed. That's because the upload speed is so much slower that it wouldn't
sound that great in megabits per second. For example, one popular carrier offers its Extreme
ADSL that has a download speed of 10Mbps and an upload speed of 512Kbps. Sounds
pretty good, doesn't? Well, what if I offered it to you with 10Mbps download and
only 0.5Mbps upload--how much would you buy then? As you can see, they are both
the same, but many people don't catch this fact.
The upload speed of ADSL is generally about ½0th its download
speed. The reason that most people buy it anyway is that they don't really care
too much about the upload speed. Most of what they do that is bandwidth
intensive is downloading, such as surfing the Internet, watching movies, and
pulling down files. Now, if they were building a website or transferring files
to an FTP server, that would be a different story altogether. However, most
people aren't doing that, so ADSL is fine for them.
For those who want a little more upload speed, a DSL service is
available in some areas that provides a balance of upload and download speed. It
is referred to as symmetric digital subscriber line
(SDSL). Typically, you won't get the fastest download speed with this
option, but that is not usually what you are after anyway. If you are
considering this option, you are one of the few who really does put large files
back onto the Web, such as when building a website or sending files to an FTP
server. The additional upload speed will save you considerable time and
headaches from watching that agonizingly slow progress indicator line, if you
know what I mean!
At the time of this writing, there were many areas that offered
SDSL at rates of up to 3Mbps but usually about 1.5Mbps. You might recognize that
as about the same as a T1. This is no coincidence since many small businesses
consider SDSL to be a less expensive option that gives them essentially what the
T1 would have, especially if they are using the bandwidth in its entirety and
not dividing the DS0s. Not all areas offer SDSL, but in many areas it can be
used to provide bandwidth acceptable to a business and still save money vs.
leasing a dedicated T1 line.
What if you want to “have your cake and eat it too”? In other
words, what if you want very high bandwidth for both upstream and downstream so
that you can watch your movies in HD and upload large files all at the same
time? In that case, you will need very high bitrate digital
subscriber line (VDSL). Currently, service providers are experimenting with
new lines that will provide more than 100Mbps (that's right, I said 100Mbps) for
both upstream and downstream simultaneously on regular telephone lines! As you
might imagine, this is still an emerging technology, and you had better be
prepared to pay for it, at least compared to what you pay now for ADSL or
SDSL.
This morning when I got to my office computer, the Internet was
down. I quickly determined that everything was connected and even tried a
restart of my cable modem, but to no avail. I called the provider who said that
they were aware of a problem in my area and would have it fixed within a couple
of hours. Hours! Without Internet? I have a business to
run, you know!
Fortunately, I also have a wireless card that fits into any USB
port on a computer and then connects to my cellular provider. Mine gives me
7.2Mbps download speed, which is fast enough for a backup line. Because of it, I
was able to conduct business rather normally this morning while my primary
provider got their issue solved.
The latest 4G technology of these cards can operate at speeds
up to 100Mbps, so these types of cards can be used for primary Internet as well
as backup lines. There are even some now that allow you to set up your own
network and allow a few other users to connect as well. There are two competing
standards, Global System for Mobile communications (GSM) and Code-Division
Multiple Access. Whether you end up using one or the other of the standards, or
even a hybrid of the two, will be determined by the cellular vendor that you
choose.
Worldwide Interoperability for Microware Access (WiMAX) is a
telecommunications protocol that is sometimes referred to as “Wi-Fi on
Steroids.” It can be used for a number of applications including broadband
connections, but it can also permit usage at much greater distances than Wi-Fi.
On a larger scale, such as for an entire community or even a small country,
WiMAX is much more cost effective. Identified by the IEEE 802.16 standard, it is
being experimented with for “last-mile” connectivity options rather than using
cable or DSL. The subscriber uses a WiMAX card that connects to their computer,
usually in a USB port. Originally, a few years ago, WiMax could deliver speeds
up to 40Mbps, but it can now deliver speeds up to 1Gbps. That's quite an
increase in such a short amount of time!
The Long Term Evolution (LTE) format was first proposed by NTT
DoCoMo of Japan and has been adopted as an international standard. At the time
of this writing, it is still in the making as far as actually being offered by
carriers. When it is released, it will offer over 100Mbps speed on wireless
links for phones, PDAs, and computers! It will be the true fourth-generation
(4G) standard, although some carriers have “jumped the gun” and are calling
enhancements to 3G a 4G standard.
High Speed Packet Access + (HSPA+) is a
wireless broadband standard that is used by some vendors for access to the
Internet on cell phones and PDAs. Also known as Evolved HSPA, it provides data
link rates of approximately 84Mbps down and 22Mbps up. Providers usually offer
some service plans at these maximum speeds and others at a slower
speed.
As I mentioned earlier, fiber-optic cable (fiber) is often used
in network backbones to provide high bandwidth for fast, reliable
communications. In addition, some companies, such as Verizon, are now offering
fiber-optic cable connections to the user's desktop. This technology can deliver
Internet speeds up to 150Mbps, but the availability is currently quite limited.
It uses an optical network terminal (ONT), which is provided by the vendor and
generally offers wired and wireless connections.
Also called Plain Old Telephone Service (POTS) or Public
Switched Telephone Network (PSTN), dialup service offers
agonizingly slow (in today's terms) 56Kbps lines.
POTS is a term that telephone company employees assigned to
those public switched lines when newer and more “sophisticated” links such as
ISDN, T1, T3, and the like began to emerge. The point was that the normal
modem-based dialup communications ran on the same lines that everyone had been
talking on for almost 100 years! What you should know about dialup are the
advantages (yes, there are some) and disadvantages of it vs. the newer
technologies that I've covered.
The main advantages of dialup are availability and cost. It is
highly available, since almost everyone has a regular telephone line (although
that is beginning to change now), and it's available at a relatively low cost
when compared to other services. As you may have guessed, the major disadvantage
of dialup lines is that they do not support the bandwidth that we need in order
to do all of the “fancy stuff” that we want to do on our computers today, such
as download movies and large files, surf multiple websites at the same time, and
hold video conferences with our peers. Still, some people in the world will hear
those screeching tones of the modem handshake when they dial up to their ISP
today.
A Passive Optical Network (PON) is a
point-to-multipoint fiber-to-premises network that allows a single fiber-optic
cable to serve multiple premises. The premises can be businesses or homes. As
technology advances, more subscribers are using fiber-optic cable to connect the
Internet. This is continuing to grow, and more types of PONs are evolving to
work with different technologies and with various providers. Many versions of
this technology are evolving, and speeds range from less than 1Gbps up to
10Gbps.
Organizations that have many locations across a wide geographic
region and that want to connect each of those locations to each other have many
options with today's communications networks. They could connect them all with
dedicated lines. This could be very expensive, though, since they would need
many dedicated lines. For example, if an organization had just five locations
that it wanted to mesh fully with dedicated lines, it would require 15 dedicated
lines.
Instead of using dedicated lines, another option available to
organizations is to use a network of switches and special routers that spans the
globe and can be connected to at any point. These lines take the information
from a computer or other host and relay it to its final destination. This type
of networking is referred to as Frame Relay because the
layer 2 frames are actually being relayed across the switches and special
routers instead of being sent on dedicated lines. This has been an effective
method of communication for many companies over the last 10 to 15 years, but it
is gradually being phased out now because of even better communication
options.
If you take a Cisco or other type of WAN-based class, you will
no doubt learn the details, and a myriad of terminology, about Frame Relay. For
the purposes of this course, you should just know that it is a method of using
common (nondedicated) lines to communicate network traffic at layer 2 (the Data
Link layer) so as to join two hosts within the same subnet. The layer 2 address
that it uses, referred to as a Data Link Connection Identifier
(DLCI), is generally assigned by the service provider such as AT&T or
Sprint. The guaranteed communication rate that the service provider agrees to is
the committed information rate (CIR). Thus, the main
reason that you might choose Frame Relay over dedicated lines is that you can
get an acceptable CIR for your connections for much less money than you would
pay to have dedicated lines for each one.
I like to joke that ATM is a technology
that allows you to take money out of your bank! Actually, ATM does not stand for
automatic teller machine (at least in this case) but for asynchronous transfer
mode. Unfortunately, knowing what the acronym means still doesn't tell you much
about what this protocol does. In essence, it's a protocol that was developed
after Ethernet, and it provides a much more efficient way of transferring data
than does Ethernet.
ATM was originally developed in the mid-1980s to be used for
voice, data, and video applications. We needed a more efficient protocol to
provide movies and sound for training and for fun. It uses a fixed-length cell
of 53 bytes, rather than the variable-length packets that are used by Ethernet.
This allows for more efficiency, since the devices never have to fragment and
reassemble large packets. The original ATM technology was already much faster
than Ethernet; it was able to transfer voice, data, and video signals at up to
500Mbps. It's now even faster and is being used by some telecommunication and
Internet providers as a backbone or core layer. ATM, like every other
technology, will eventually be replaced by faster successors.
So, just how do you decide which one of these types of WAN
connections is best for your situation? Well, a good place to start is to
compare the properties of each type of connection against the needs of the
organization in which they will be used. Some of the many properties that you
should consider include whether the solution is circuit switched or packet
switched, its speed relative to other solutions, the media that is used, and the
distance that media can carry the data. In the following sections, I'll discuss
each of these important properties of WAN communication.
The properties of WAN technology types are a description of the
events that happen and/or the attributes of the communication. In the case of
circuit switch, we are back to our dialup lines. Once you establish a connection
on a circuit-switched network, the entire conversation or line of data traffic
is sent on those same physical connections until you terminate the connection.
If you were to establish a new connection to the same place, you would likely
get a very different set of connections that would also “complete your
call.”
Packet switch networks are very different from circuit switch
ones in that each data packet might take a different route to its final
destination during the same transmission. The original packet switch network,
called X.25, was developed to overcome the challenge of sending reliable
communications through an inherently unreliable medium. The unreliable medium at
that time was, you guessed it, Ma Bell!
A computer modem's “screeching” must have sounded very
different from those original switches than the voice of Alexander Graham Bell
saying, “Come here, Watson, I need you.” If one switch weren't getting the job
done, X.25 would just take a different path automatically. However, because of
the extensive error checking built into X.25, it was inherently very
reliable…and very slow!
The newer packet-switched networks include some that I've
already discussed in this chapter such as Frame Relay. These networks use
sophisticated virtual circuits to avoid errors and thereby improve efficiency
and data throughput. They are connection-oriented now, so data doesn't move from
one location until it's completely cleared to “land” in the next. This makes for
fast and reliable data flow.
Speed is a little deceptive in that it's a factor of available
bandwidth and throughput. In other words, just because you have a T1 doesn't
mean you have a fast connection. It depends on how many others are using it and
on what you need to do with it. Also, your speed through a network will largely
be determined by the available bandwidth on the slowest link that you
encounter.
Let's say you were to take a trip all the way across the United
States that was mostly on the expressways but you had to climb up and down
winding mountain overpasses for the last 500 miles. That last 500 miles would
likely take the bulk of your time on the trip, even though it wasn't the
greatest distance. The same thing happens when user data comes from one location
to another through a very fast backbone: they may have a speed of 10Gbps on the
backbone, but as soon as they come into the building and hit your 100Mbps Fast
Ethernet switch, their relative speed for the whole connection will be, well,
100Mbps. Maybe there is more to this speed thing than first meets the
eye?
This looks like a no-brainer at first glance. Transmission
media is the stuff on which the communication is carried. This media could be
copper, fiber, or even wireless. The media you choose will depend on your
bandwidth needs, security needs, EMI concerns, and the distance you need to send
the data before it hits another switch or router. Since I'm talking about the
WAN environment, the service provider will have already made these decisions. I
will categorize many types of transmission media in the next sections of this
chapter.
Distance is “how far” data is sent. I don't think that's on the
exam, but the relative
distance of communications on various communications media might “creep in
there.” Generally speaking, fiber optics are capable of much greater distances
because their attenuation rate is nowhere near as high as that of copper
connections. This is because the signal being sent is light rather than
electricity. In regard to WAN connectivity, the good news is that it's not your
problem. The service provider should give you the right links to get you from
point A to point B in the most efficient and effective manner. Later in this
chapter, I will categorize many types of transmission media in regard to
distance and speed.
Be able to categorize types of WAN
technologies Be able to differentiate between Frame-Relay, T1, T3, ADSL,
SDSL, VDSL, cable, satellite, OC-x, wireless, ATM, SONET, ISDN, and dialup
technologies. Know the major advantages and disadvantages of each and where they
are likely to be used.
Be able to categorize WAN properties
Understand the difference between circuit switch and packet switch. You should
be able to categorize WAN properties with regard to speed, transmission media,
and distance limitations.
Today's organizations use many different types of WAN
connections. Your decision to use one connection type over another will largely
depend on the properties each connection type offers and the advantages it may
provide your organization. In the following sections, I'll discuss the main WAN
connection types and their properties.
Types
Most WAN connections are provided by a communications company
referred to as a service provider. One of the main
differences between a WAN and a LAN is that you generally don't own all the
connections on a WAN, unless you are the service provider. Because of this, the
types of network connections you will encounter in a WAN environment are very
different from the connections you are used to in a LAN environment. Most
network connection types have evolved over time, offering the right properties
to connect a company's computers based on the needs of the company for its
connections. I'll discuss WAN connection types such as Frame Relay, T1, T3, DSL,
cable modem, ATM, SONET, wireless, and ISDN.T1/E1
Let's say your company has one location in New York and another
in San Diego. For your communications, you would like to have a cable that
connects the two locations. You could get a very big truck and a whole lot of
cable and just start driving cross-country, carefully spooling out the cable and
telling everyone not to bother it. Of course, it still wouldn't work when you
got to the end because of the attenuation of copper wire, but that's another
story!This ridiculous example will help you see what service
providers have done for you. In essence, they have already rolled out that
cable, but they require payment to let you connect to both ends of it. Once you
do, that's your connection, and nobody else is on it. The service provider will
generally “condition” the line from time to time, testing it and making sure you
are getting what you are paying for, but you will be the only one authorized to
use it to communicate. Sometimes this division is only logical, but often it is
physical as well; in other words, it's your wire and only your wire at many
points. Cool, huh?The most common of these types of connections in the United
States has been the T1. In Europe, they use a very similar
connection called an E1. A T1 provides for 1.54Mbps of
dedicated bandwidth for the customer. This bandwidth can be used in total or
divided up into as many as 24 channels, called DS0s, which are each 64Kbps. An
E1 is very similar but offers 32 DS0s instead of 24 for a total of 2.048Mbps.
This is accomplished using a device called a CSU/DSU. This gives the customer
(you) many options in regard to the dispensation of the bandwidth to network
resources. The cost of a T1 or an E1 varies based on the connection points you
choose, but it's safe to say you would pay between $500 and $1,000 a month for
one T1 line in most locations today. This is why an organization will consider
other options before spending the money on a dedicated T1 connection. I should
also mention that a T1 is not (by far) the most expensive connection type you
can choose.T3/E3/DS3
Caution, the next couple of options are big bucks! What if you
wanted a line like a T1 but much, much larger? Some large companies require
high-bandwidth, dedicated connections from one office or data center to another.
One way to accomplish this is by using a T3 line in the
United States or an E3 line in Europe and much of the rest
of the world. This type of line provides a tremendous amount of usable bandwidth
that can be divided to fit an organization's needs. A T3 provides for 672 DS0s,
or the equivalent of 28 T1s or 44.736Mbps! It is sometimes also referred to as a
DS3. An E3 provides for the equivalent of 512 DS0s, or
approximately 17 E1s or 34.368Mbps.OC-x
If you thought those were fast, you ain't seen nothin' yet! In
the term OC-x, the OC stands for “optical carrier,” and the x indicates the
relative speed of the link. Well, the x just keeps getting bigger and bigger.
The original speed of an OC trunk was about 50Mbps, and it was called an OC-1.
OC-3 quickly followed with 150Mbps. The standard at the time of this writing is
OC-3072, which offers a mind-blowing data rate of 160Gbps--that's with a G! As
you might imagine, only very large companies have the need (or the money) for
these options.SONET
How about a protocol that will push data at 150Gbps over fiber
links and has to be controlled using atomic clocks? That's what Synchronous
Optical Networking (SONET) can do. It's especially useful for networks that span
multiple geographic regions because the atomic clock mechanism in it keeps
everyone on the same exact millisecond. As you can imagine, this protocol
requires expensive equipment and expertise. It is typically used by large
communication providers and very large corporations as a transfer mechanism or
backbone for data traffic. In regard to speeds, SONET generally provides an
STS-1 link of 50Mbps bandwidth and multiples thereof, whereas STS-3 would
provide about 150Mbps bandwidth.SDH
Synchronous Digital Hierarchy (SDH) is a standard that is very
similar to SONET. It transfers data over optical fiber using laser light or LED.
Its speed and capability are similar to SONET, and its timing is controlled
using atomic clocks. SDH was originally defined by the European
Telecommunications Standards Institute, whereas SONET was originally defined by
the American National Standards Institute. In regard to speeds, SDH defines an
STM-0 at 50Mbps bandwidth.DWDM
Dense Wavelength Division Multiplexing (DWDM) is an optical
technology used to increase bandwidth even further over existing fiber-optic
backbones. It works by combining and transmitting multiple signals
simultaneously at different wavelengths on the same fiber. One fiber is
essentially turned into multiple “virtual fibers.” Currently, because of DWDM,
single fibers have been able to transmit data at speeds up to 400Gbps.Satellite
Earlier, when I was talking about “coaxial cable that can
connect you to the Internet,” some of you might have been thinking “What cable?
We live out in the countryside and don't have those cables out here!” If that's
the case, then your best (and maybe only) option is a satellite hookup. You may
also decide to use a satellite hookup because it is the most economical or
dependable service in your area. In either case, you will need a dish antenna
and a professional installer or instructions on how to find their satellite with
your dish antenna. It always makes me laugh when I hear someone say that they
put the satellite in the backyard to get the TV signal. That's not the
“satellite.” The satellite is in geostationary orbit high above the earth, and
your little antenna is just going to pick up on its signal from space.Once you do that, you can then use that signal as a download
from the Internet. Now uploading is a bit trickier, since you probably won't
have a high-powered transponder. Many satellite communications companies have
provided the upload through your regular telephone line dialup connections. Some
now provide a DSL line to give you more bandwidth for uploads. Others advertise
a two-way satellite system that actually does send some signal back to the
satellite from the antenna. These are typically more expensive and harder to
install. Satellite communications companies offer data rates that rival those of
their biggest competitors, cable and DSL.ISDN
This is one that I can't believe is still in the exam objectives! It's been so
long since we used ISDN that it should just go away, and we shouldn't need to
know the details anymore. Oh well, since it's listed on the objectives, I'll go
over the most important aspects regarding ISDN BRI and ISDN PRI that you might
need to know for the test…and very unlikely for real
life.ISDN BRI is a layer 2 protocol that allows for two
communication channels and one control channel. The communication channels are
referred to as B (bearer) channels, and the control channel is referred to as a
D (delta) channel. Each of the B channels can carry up to 64Kbps of data (that
used to be a lot), and the control channel can use 16Kbps for data control. ISDN
BRI is sometimes referred to as 2B + D, but this is actually misleading because
the B channels are really the only usable bandwidth for data. Thus, an ISDN line
can carry a whopping 128Kbps of data! In other words, one T1 is the rough
equivalent of 12 ISDN BRI lines. ISDN also employs all kinds of telephone
company terminology that identifies the reference points and the devices, but
the chance of you having to know that for this exam (or for real life
anymore) is so remote that it hardly bears mentioning.ISDN PRI came out a little later, and it's a very different
story than ISDN BRI. An ISDN PRI link is almost the same as a T1 line in regard
to its capacity to carry data. It consists of 23 B channels (each with 64Kbps)
and 1 D channel (also with 64Kbps). A little quick math should tell you that a
single ISDN link will carry 1,472Kbps, which, as you can see, is very close to a
T1.So, why would a company choose an ISDN PRI link over a T1?
Actually, it could come down to the availability in an area, the cost, the type
of equipment that the company already owns, and the business rationale for the
link. Both ISDN PRI and T1 links offer 23 DS0s for a customer's actual bandwidth
use. (That's the “dirty little secret” about T1s--you don't get all 1.544Mbps
for your use, but you should still know that number for the exam.) Some equipment has
ISDN PRI interfaces built in, which allow flexible control of the 23 B channels,
so a company can use them for special needs such as video conferences or
network-based meetings. ISDN PRI also employs a myriad of telephone company
jargon that, thankfully, you will not need to know. Just know that it's 23 B + D
and that all channels are 64Kbps DS0s, and you will be fine.Cable
Cable companies have jumped on the bandwagon and now offer you
a path to the Internet that begins with connecting your computer to a special
cable modem. That modem is configured by the cable company to be recognized by
its central office, also called the headend of the cable
company. From there, the cable company becomes your Internet service provider
(ISP), connecting you to the Internet.Many small businesses and home users have chosen this option
for their Internet connection. The advantages include the fact that most cable
companies can provide tremendous bandwidth (10Mbps and faster) downstream and
“acceptable” bandwidth (over 1Mbps) upstream. For most users and small
businesses, this is all they really need. A potential disadvantage is that you
will share bandwidth with others who are in your immediate service area. This
means that at peak times your performance could become degraded. This is not the
case with the next technology that I will discuss.In the early 1990s, when ISDN BRI was just coming out, I was
working at Sprint in Florida. We used to joke that ISDN stood for either “It
Still Does Nothing” or “I Still Don't kNow.” Sprint wanted about 90 bucks a
month for it so that you could combine your Internet service with your telephone
services. The problem was that most people at that time were saying “What's an
Internet?” You see, the Internet at that point was accessible only by text
commands and usually only through a university or a large corporation. It wasn't
all that exciting then, except for a few of us “geeks.”By the mid-1990s, though, with the ushering in of the World
Wide Web and hypertext browsing, everything had changed. We sold a lot of ISDN
BRI during that time until the newer technologies of ADSL and cable modem
Internet took its place. Then we still sold it as backup lines to companies that
were leasing a T1 or a T3. It was no longer an expensive add-on by that time.
Now ISDN BRI is all but gone, but you still need to know the historical
information covered in this chapter for the exam.DSL
Another inexpensive option for small companies and home users,
which provides considerable bandwidth for an economical rate, is a digital subscriber line (DSL). The most common of these is
the asymmetric digital subscriber line (ADSL). The reason
this type of connection can be economical is that the lines it uses are already
in place. They are your regular telephone lines. The service provider uses the
regular telephone lines and special equipment that multiplexes the signal to
provide tremendous bandwidth over that which dialup lines provide.Now, as you may know, asymmetric means “not the same on both
sides” or “not balanced.” So, what is not balanced about an ADSL line? The
bandwidth is not balanced; in fact, it's not even close. You may have noticed
that service providers advertise ADSL using megabits per second for download speed but
kilobits per second for
upload speed. That's because the upload speed is so much slower that it wouldn't
sound that great in megabits per second. For example, one popular carrier offers its Extreme
ADSL that has a download speed of 10Mbps and an upload speed of 512Kbps. Sounds
pretty good, doesn't? Well, what if I offered it to you with 10Mbps download and
only 0.5Mbps upload--how much would you buy then? As you can see, they are both
the same, but many people don't catch this fact.The upload speed of ADSL is generally about ½0th its download
speed. The reason that most people buy it anyway is that they don't really care
too much about the upload speed. Most of what they do that is bandwidth
intensive is downloading, such as surfing the Internet, watching movies, and
pulling down files. Now, if they were building a website or transferring files
to an FTP server, that would be a different story altogether. However, most
people aren't doing that, so ADSL is fine for them.For those who want a little more upload speed, a DSL service is
available in some areas that provides a balance of upload and download speed. It
is referred to as symmetric digital subscriber line
(SDSL). Typically, you won't get the fastest download speed with this
option, but that is not usually what you are after anyway. If you are
considering this option, you are one of the few who really does put large files
back onto the Web, such as when building a website or sending files to an FTP
server. The additional upload speed will save you considerable time and
headaches from watching that agonizingly slow progress indicator line, if you
know what I mean!At the time of this writing, there were many areas that offered
SDSL at rates of up to 3Mbps but usually about 1.5Mbps. You might recognize that
as about the same as a T1. This is no coincidence since many small businesses
consider SDSL to be a less expensive option that gives them essentially what the
T1 would have, especially if they are using the bandwidth in its entirety and
not dividing the DS0s. Not all areas offer SDSL, but in many areas it can be
used to provide bandwidth acceptable to a business and still save money vs.
leasing a dedicated T1 line.What if you want to “have your cake and eat it too”? In other
words, what if you want very high bandwidth for both upstream and downstream so
that you can watch your movies in HD and upload large files all at the same
time? In that case, you will need very high bitrate digital
subscriber line (VDSL). Currently, service providers are experimenting with
new lines that will provide more than 100Mbps (that's right, I said 100Mbps) for
both upstream and downstream simultaneously on regular telephone lines! As you
might imagine, this is still an emerging technology, and you had better be
prepared to pay for it, at least compared to what you pay now for ADSL or
SDSL.Cellular
This morning when I got to my office computer, the Internet was
down. I quickly determined that everything was connected and even tried a
restart of my cable modem, but to no avail. I called the provider who said that
they were aware of a problem in my area and would have it fixed within a couple
of hours. Hours! Without Internet? I have a business to
run, you know!Fortunately, I also have a wireless card that fits into any USB
port on a computer and then connects to my cellular provider. Mine gives me
7.2Mbps download speed, which is fast enough for a backup line. Because of it, I
was able to conduct business rather normally this morning while my primary
provider got their issue solved.The latest 4G technology of these cards can operate at speeds
up to 100Mbps, so these types of cards can be used for primary Internet as well
as backup lines. There are even some now that allow you to set up your own
network and allow a few other users to connect as well. There are two competing
standards, Global System for Mobile communications (GSM) and Code-Division
Multiple Access. Whether you end up using one or the other of the standards, or
even a hybrid of the two, will be determined by the cellular vendor that you
choose.WiMAX
Worldwide Interoperability for Microware Access (WiMAX) is a
telecommunications protocol that is sometimes referred to as “Wi-Fi on
Steroids.” It can be used for a number of applications including broadband
connections, but it can also permit usage at much greater distances than Wi-Fi.
On a larger scale, such as for an entire community or even a small country,
WiMAX is much more cost effective. Identified by the IEEE 802.16 standard, it is
being experimented with for “last-mile” connectivity options rather than using
cable or DSL. The subscriber uses a WiMAX card that connects to their computer,
usually in a USB port. Originally, a few years ago, WiMax could deliver speeds
up to 40Mbps, but it can now deliver speeds up to 1Gbps. That's quite an
increase in such a short amount of time!LTE
The Long Term Evolution (LTE) format was first proposed by NTT
DoCoMo of Japan and has been adopted as an international standard. At the time
of this writing, it is still in the making as far as actually being offered by
carriers. When it is released, it will offer over 100Mbps speed on wireless
links for phones, PDAs, and computers! It will be the true fourth-generation
(4G) standard, although some carriers have “jumped the gun” and are calling
enhancements to 3G a 4G standard.HSPA+
High Speed Packet Access + (HSPA+) is a
wireless broadband standard that is used by some vendors for access to the
Internet on cell phones and PDAs. Also known as Evolved HSPA, it provides data
link rates of approximately 84Mbps down and 22Mbps up. Providers usually offer
some service plans at these maximum speeds and others at a slower
speed.Fiber
As I mentioned earlier, fiber-optic cable (fiber) is often used
in network backbones to provide high bandwidth for fast, reliable
communications. In addition, some companies, such as Verizon, are now offering
fiber-optic cable connections to the user's desktop. This technology can deliver
Internet speeds up to 150Mbps, but the availability is currently quite limited.
It uses an optical network terminal (ONT), which is provided by the vendor and
generally offers wired and wireless connections.Dialup
Also called Plain Old Telephone Service (POTS) or Public
Switched Telephone Network (PSTN), dialup service offers
agonizingly slow (in today's terms) 56Kbps lines.POTS is a term that telephone company employees assigned to
those public switched lines when newer and more “sophisticated” links such as
ISDN, T1, T3, and the like began to emerge. The point was that the normal
modem-based dialup communications ran on the same lines that everyone had been
talking on for almost 100 years! What you should know about dialup are the
advantages (yes, there are some) and disadvantages of it vs. the newer
technologies that I've covered.The main advantages of dialup are availability and cost. It is
highly available, since almost everyone has a regular telephone line (although
that is beginning to change now), and it's available at a relatively low cost
when compared to other services. As you may have guessed, the major disadvantage
of dialup lines is that they do not support the bandwidth that we need in order
to do all of the “fancy stuff” that we want to do on our computers today, such
as download movies and large files, surf multiple websites at the same time, and
hold video conferences with our peers. Still, some people in the world will hear
those screeching tones of the modem handshake when they dial up to their ISP
today.PON
A Passive Optical Network (PON) is a
point-to-multipoint fiber-to-premises network that allows a single fiber-optic
cable to serve multiple premises. The premises can be businesses or homes. As
technology advances, more subscribers are using fiber-optic cable to connect the
Internet. This is continuing to grow, and more types of PONs are evolving to
work with different technologies and with various providers. Many versions of
this technology are evolving, and speeds range from less than 1Gbps up to
10Gbps.Frame Relay
Organizations that have many locations across a wide geographic
region and that want to connect each of those locations to each other have many
options with today's communications networks. They could connect them all with
dedicated lines. This could be very expensive, though, since they would need
many dedicated lines. For example, if an organization had just five locations
that it wanted to mesh fully with dedicated lines, it would require 15 dedicated
lines.Instead of using dedicated lines, another option available to
organizations is to use a network of switches and special routers that spans the
globe and can be connected to at any point. These lines take the information
from a computer or other host and relay it to its final destination. This type
of networking is referred to as Frame Relay because the
layer 2 frames are actually being relayed across the switches and special
routers instead of being sent on dedicated lines. This has been an effective
method of communication for many companies over the last 10 to 15 years, but it
is gradually being phased out now because of even better communication
options.If you take a Cisco or other type of WAN-based class, you will
no doubt learn the details, and a myriad of terminology, about Frame Relay. For
the purposes of this course, you should just know that it is a method of using
common (nondedicated) lines to communicate network traffic at layer 2 (the Data
Link layer) so as to join two hosts within the same subnet. The layer 2 address
that it uses, referred to as a Data Link Connection Identifier
(DLCI), is generally assigned by the service provider such as AT&T or
Sprint. The guaranteed communication rate that the service provider agrees to is
the committed information rate (CIR). Thus, the main
reason that you might choose Frame Relay over dedicated lines is that you can
get an acceptable CIR for your connections for much less money than you would
pay to have dedicated lines for each one.ATMs
I like to joke that ATM is a technology
that allows you to take money out of your bank! Actually, ATM does not stand for
automatic teller machine (at least in this case) but for asynchronous transfer
mode. Unfortunately, knowing what the acronym means still doesn't tell you much
about what this protocol does. In essence, it's a protocol that was developed
after Ethernet, and it provides a much more efficient way of transferring data
than does Ethernet.ATM was originally developed in the mid-1980s to be used for
voice, data, and video applications. We needed a more efficient protocol to
provide movies and sound for training and for fun. It uses a fixed-length cell
of 53 bytes, rather than the variable-length packets that are used by Ethernet.
This allows for more efficiency, since the devices never have to fragment and
reassemble large packets. The original ATM technology was already much faster
than Ethernet; it was able to transfer voice, data, and video signals at up to
500Mbps. It's now even faster and is being used by some telecommunication and
Internet providers as a backbone or core layer. ATM, like every other
technology, will eventually be replaced by faster successors.
Properties
So, just how do you decide which one of these types of WAN
connections is best for your situation? Well, a good place to start is to
compare the properties of each type of connection against the needs of the
organization in which they will be used. Some of the many properties that you
should consider include whether the solution is circuit switched or packet
switched, its speed relative to other solutions, the media that is used, and the
distance that media can carry the data. In the following sections, I'll discuss
each of these important properties of WAN communication.Circuit switch
The properties of WAN technology types are a description of the
events that happen and/or the attributes of the communication. In the case of
circuit switch, we are back to our dialup lines. Once you establish a connection
on a circuit-switched network, the entire conversation or line of data traffic
is sent on those same physical connections until you terminate the connection.
If you were to establish a new connection to the same place, you would likely
get a very different set of connections that would also “complete your
call.”Packet switch
Packet switch networks are very different from circuit switch
ones in that each data packet might take a different route to its final
destination during the same transmission. The original packet switch network,
called X.25, was developed to overcome the challenge of sending reliable
communications through an inherently unreliable medium. The unreliable medium at
that time was, you guessed it, Ma Bell!A computer modem's “screeching” must have sounded very
different from those original switches than the voice of Alexander Graham Bell
saying, “Come here, Watson, I need you.” If one switch weren't getting the job
done, X.25 would just take a different path automatically. However, because of
the extensive error checking built into X.25, it was inherently very
reliable…and very slow!The newer packet-switched networks include some that I've
already discussed in this chapter such as Frame Relay. These networks use
sophisticated virtual circuits to avoid errors and thereby improve efficiency
and data throughput. They are connection-oriented now, so data doesn't move from
one location until it's completely cleared to “land” in the next. This makes for
fast and reliable data flow.Speed
Speed is a little deceptive in that it's a factor of available
bandwidth and throughput. In other words, just because you have a T1 doesn't
mean you have a fast connection. It depends on how many others are using it and
on what you need to do with it. Also, your speed through a network will largely
be determined by the available bandwidth on the slowest link that you
encounter.Let's say you were to take a trip all the way across the United
States that was mostly on the expressways but you had to climb up and down
winding mountain overpasses for the last 500 miles. That last 500 miles would
likely take the bulk of your time on the trip, even though it wasn't the
greatest distance. The same thing happens when user data comes from one location
to another through a very fast backbone: they may have a speed of 10Gbps on the
backbone, but as soon as they come into the building and hit your 100Mbps Fast
Ethernet switch, their relative speed for the whole connection will be, well,
100Mbps. Maybe there is more to this speed thing than first meets the
eye?Transmission media
This looks like a no-brainer at first glance. Transmission
media is the stuff on which the communication is carried. This media could be
copper, fiber, or even wireless. The media you choose will depend on your
bandwidth needs, security needs, EMI concerns, and the distance you need to send
the data before it hits another switch or router. Since I'm talking about the
WAN environment, the service provider will have already made these decisions. I
will categorize many types of transmission media in the next sections of this
chapter.Distance
Distance is “how far” data is sent. I don't think that's on the
exam, but the relative
distance of communications on various communications media might “creep in
there.” Generally speaking, fiber optics are capable of much greater distances
because their attenuation rate is nowhere near as high as that of copper
connections. This is because the signal being sent is light rather than
electricity. In regard to WAN connectivity, the good news is that it's not your
problem. The service provider should give you the right links to get you from
point A to point B in the most efficient and effective manner. Later in this
chapter, I will categorize many types of transmission media in regard to
distance and speed.
Exam Essentials
Be able to categorize types of WAN
technologies Be able to differentiate between Frame-Relay, T1, T3, ADSL,
SDSL, VDSL, cable, satellite, OC-x, wireless, ATM, SONET, ISDN, and dialup
technologies. Know the major advantages and disadvantages of each and where they
are likely to be used.Be able to categorize WAN properties
Understand the difference between circuit switch and packet switch. You should
be able to categorize WAN properties with regard to speed, transmission media,
and distance limitations.