Basic Communication Technology

CHAPTER 3
BASIC DATA COMMUNICATION

TECHNOLOGY
Concepts Reinforced
Top-down model I-P-O Model
Modulation techniques Protocols and compatibility
Transmission services Multiplexing
Concepts Introduced
Communications media Cable modem technologies
Wireless point-to-point Serial technologies
technologies Internet service providers
Dial-up modem technologies DSL technologies
OBJECTIVES
After mastering the material in this chapter you should:
1. Understand the types of communication media available and the limitations

and applications of each.
2. Be able to compare and contrast basic point-to-point communication tech-
nologies such as RS-232, USB, IEEE-1394/Firewire, and Centronics parallel.
3. Understand the applications and limitations of wireless point-to-point com-
munication technologies, including IrDA and Bluetooth.
4. Understand the role of an Internet service provider.
5. Understand modem operation, comparative modem features, the impor-
tance of modem standards, and the cost/benefit analysis of various
modem purchases.
6. Understand the architecture and operation of digital subscriber line (DSL)
technologies.
7. Understand the architecture and operation of cable modem technologies.
8. Be able to analyze and select the best Internet access technology for a
given location.
72 Chapter Three/Basic Data Communication Technology
■ INTRODUCTION
This chapter begins an in-depth analysis of the OSI Network Reference Model layers
that is carried forward through chapter 8. As detailed in chapter 1, the OSI Network
Reference Model can easily be mapped onto other protocol models such as the
TCP/IP model. This mapping will be detailed throughout the coming chapters.
This chapter introduces communication media, basic point-to-point data com-
munications technologies, and Internet access technologies. It takes the basic con-
cepts introduced in chapter 2 and maps them to real-world technologies that a
typical student interacts with on a daily basis. In addition to technical details,
emphasis is placed on the application and business implications of the technolo-
gies introduced.
■ PHYSICAL LAYER
The first layer of the OSI Network Reference Model, the physical layer, is responsible
for the establishment, maintenance, and termination of physical connections
between communicating devices. The physical layer transmits and receives a stream
of bits. There is no data recognition at the physical layer. The physical layer is
mapped onto the TCP/IP model as the bottom part of the network access layer that
also contains the OSI datalink layer.
Specifically, physical layer operation is controlled by protocols that define the
electrical, mechanical, and procedural specifications for data transmission. Strictly
speaking, the physical layer does not define the specifications for media and connec-
tors. These specifications are sometimes referred to as belonging to Layer zero of the
model, as they underlie layer one.
■ COMMUNICATIONS MEDIA
Although communications media are technically not part of the OSI Network Refer-
ence or TCP/IP models, their importance cannot be understated. Without some sort
of connection between two devices they will not be able to communicate; regardless
of the hardware and protocols used. In this way media is analogous to a highway; it
does not do you any good to have a Ferrari if you need to travel across a swamp that
has no roads.
When two devices that have been working flawlessly cease to operate, there is a
very good chance that the problem can be directly attributed to problems with the
media connecting the devices. Perhaps the first mantra of network troubleshooting
should be, “Check the physical layer.”
There are many different types of data communications media. This section
details commonly used media, their characteristics, and typical application.
Not Twisted Pair
The type of phone wire installed in most homes built prior to the mid-1990s consists
of a tan plastic jacket containing four untwisted wires: red, yellow, green, and black.
This cable is typically referred to as four conductor station wire or RYGB. While
Communications Media 73
capable of carrying an analog telephone call, this type of wire is not suitable for data
transmission and should not be confused with unshielded twisted pair (UTP).
Another popular type of phone wiring is flat gray modular cable, also known as
gray satin or silver satin. Inside this flat gray jacket are two, four, six, or eight wires,
which are crimped into either RJ-11 (4 wire), RJ-12 (6 wire), or RJ-45 plugs (8-wire)
using a specialized crimping tool. Premises phone wiring as well as phones; crimp
tools, RJ plugs, and flat gray modular wire are attainable at nearly any hardware or
department store.
Flat gray modular wire is not the same as twisted pair and is suitable only for
carrying data at fairly slow speeds over short distances. For instance, this type of
cable is often used between a PC or workstation and a nearby RJ-11 jack for access to
the telephone system or to carry RS-232 serial communication signals.
Unshielded Twisted Pair
Twisted-pair wiring consists of one or more pairs of insulated copper wire twisted
around each other at varying lengths ranging from 2 to 12 twists per foot. The twist-
ing is used as a mechanism to reduce interference between pairs and from outside
sources such as electric motors and fluorescent lights that can cause data errors and
necessitate retransmission. These individually twisted pairs are then grouped
together and covered with a plastic or vinyl jacket, or sheath. No additional shielding
is added before the pairs are wrapped in the plastic jacket. Thus, the completed prod-
uct is known as unshielded twisted pair or UTP. The most common numbers of
pairs combined to form the unshielded twisted pair cables are 2, 3, 4, and 25 pairs of
twisted copper wire..
All UTP cables are not created equal. One of the common appeals of UTP is that it
is often already installed in modern buildings for the purpose of carrying voice con-
versations from telephone handsets to a voice PBX. Most often, when the twisted-pair
wiring for the voice PBX was installed, extra pairs were wired to each office location.
Some people jump to the conclusion that they don’t need to invest in any new wiring
to carry data transmission throughout their building–they can just use the extra pairs
of the existing UTP wiring. The problem lies in the fact that there are six different cate-
gories of UTP as specified by EIA/TIA 568 (Electronics Industry Association/Telecom-
munications Industry Association). EIA/TIA 568 also specifies the following:
• The topology, cable types, and connector types to be used in EIA/TIA 568
compliant wiring schemes
• The minimum performance specifications for cabling, connectors and com-
ponents—such as wall plates, punch down blocks, and patch panels to be
used in an EIA/TIA 568 compliant installation
Although category 1 UTP, otherwise known as voice-grade, need only carry

voice conversations with reasonable clarity, categories 3 to 6 (data-grade) cable must
meet certain predefined electrical characteristics that assure transmission quality and
speed. Before assuming that the UTP in a building is adequate for data transmission,
have its transmission characteristics tested to be certain that these characteristics
meet the listed data-grade UTP specifications. Figure 3-1 summarizes the specifica-
tions for categories 1 to 6 UTP. It is important to note that a higher category of cable
can be used for applications that require a lower cable category.
UTP Maximum Data Attenuation

Category Speed /NEXT limit Applications
Cat 1 UTP < 1Mbps Telephone lines
Cat 2 UTP 4 Mbps 4 MHz 4 Mbps token ring over UTP
Cat 3 UTP 16 Mbps 16 MHz 10BaseT Ethernet. Tested for attenuation and near-end
crosstalk up to 16 MHz.
Cat 4 UTP 20 Mbps 20 MHz 16 Mbps token ring over UTP. Tested for attenuation and
near-end crosstalk up to 20 MHz.
Cat 5 UTP 100 Mbps (2 pair) 100 MHz 100BaseT (fast) Ethernet, 155 Mbps ATM, gigabit Ethernet
1 Gbps (4 pair)
Cat 5e UTP 100 Mbps (2 pair) 100 MHz 100BaseT (fast) Ethernet, 155 Mbps ATM, gigabit Ethernet
1 Gbps (4 pair) Category 5e cable has a tighter quality control standard than
standard cat 5 cable.
Cat 6 UTP 2.5 Gbps (2 pair) 200 MHz None that require cat 6 at the time of this writing. The IEEE is
potentially up to working on a copper 10 Gbps Ethernet standard that would
10 Gbps (4 pair) require cat 6 if released.
Figure 3-1 Unshielded Twisted Pair Specifications
Wire thickness is measured by gauge and represented with the unit AWG
(American Wire Gauge). The higher the gauge number, the thinner the wire. UTP
wiring of different categories must meet specifications for resistance to different
forces that interfere with signal strength. Two of the more common sources of inter-
ference or loss of signal strength are as follows:
• Attenuation is the decrease in the power of signal over a distance in a partic-

ular type of wire or media.
• Near-End Crosstalk (NExT) is signal interference caused by a strong signal
on one-pair (transmitting) overpowering a weaker signal on an adjacent pair
(receiving). Near End Crosstalk and Attenuation to Crosstalk Ratio (ACR) are
both measured in dB or decibels. A decibel is a logarithmic rather than linear
measurement of the ratio between two powers, often a data signal and some
type of noise or interference.
BEYOND CAT 6
Although no official category 7 cable has become standardized, media vendors are
attempting to develop cable that is capable of carrying data at frequencies of up to
600 MHz. Some such attempts are not truly unshielded twisted pair but rather are
FTP or foil-twisted-pair cable that is more closely related to shielded twisted pair.
Practical Advice Buyers must be wary of so-called category 7 cable by focusing on whether the cable
and Information is truly UTP and whether it is a specified EIA/TIA standard.
Anixter, a major cabling manufacturer, has proposed a program to define perfor-
mance characteristics for cabling tested beyond the 100 MHz required for cat 5. Since
higher-speed network architectures such as fast Ethernet and gigabit Ethernet
require four pair of UTP to be transmitting power simultaneously, it was determined
Highest Test Required Frequency in MHz for Attenuation to Crosstalk Ratio

Level Frequency (ACR) at 10dB (Powersum Bandwidth)
5 200 MHz 80 MHz
6 350 MHz 100 MHz
7 400 MHz 160 MHz
Figure 3-2 UTP Level 5, 6, and 7 Performance Specifications
that crosstalk should be measured by taking into account the crosstalk influence
from all pairs in the cable (whether 4-pair or 25-pair) rather than just crosstalk
between adjacent pairs, or pair-to-pair, as had been required for cat 5 certification.
This type of crosstalk test is called Powersum Crosstalk. Key performance specifica-
tions for Level 5, 6, 7 are detailed in Figure 3-2.
Regardless of the final standard for category 7 cable, it is likely that the standard
will not support the venerable RJ-45 connector. The design of the RJ-45 simply pro-
motes too much crosstalk to effectively support the higher data speeds associated
with category 7. However, changing the cable could spell the end of the category 7
project. If the connector changes, then it would not be possible to use category 7 cable
and connectors with existing data communication technologies such as fast and giga-
bit Ethernet. Without this backward compatibility, many buyers might simply
choose to migrate their cable plants directly to fiber-optic cable to take advantage of
its even higher bandwidth.
COMMON UTP INSTALLATION MISTAKES

Strict adherence to EIA/TIA 568 installation standards is essential to successful
transmission at 100 Mbps or higher over UTP cabling. Because a less-than-perfect
installation will probably transport 10 Mbps traffic without any problem, issues with
noncompliant installations may not surface until upgrades to higher-speed network
Practical Advice architectures are attempted. The most common installation mistakes are:
and Information
• Untwisting the UTP wire more than the maximum 13 mm in order to secure
the UTP to wall plates or punch-down blocks. Exceeding the maximum bend
radius specified for UTP. Overbending the wire increases crosstalk between
stretched pairs of wires.
• Bundling the groups of UTP together too tightly with cable ties. Excessively
pinching the UTP together increases crosstalk between pairs.
STP-Shielded Twisted Pair
Data transmission characteristics, and therefore the data transmission speed, can be
improved by adding shielding around each individual pair and the entire group of
twisted pairs. This shielding may be a metallic foil or copper braid. The function of
the shield is simple. It “shields” the individual twisted pairs as well as the entire
cable from either EMI (Electromagnetic Interference) or RFI (Radio Frequency Inter-
ference). However, installation of shielded twisted pair can be tricky.
The shielding is metal and therefore a conductor. Often, the shielding is termi-
nated in a drain wire that must be properly grounded on one end of the cable. It is
critical that the shield be grounded on only one end or a ground loop can occur. In a
ground loop a small current flows across the shield from the ground connector at one
end of the cable to the ground connector at the other end of the cable. This current
can actually decrease performance by inducing noise directly into the signal wires—
quite the opposite of why the shielding was added.
Shielded twisted pair is becoming much less common. STP was commonly spec-
ified for token ring installations. However, recent technologies such as high-speed
Ethernet, ATM, and other high-speed network architectures have specified category
5+ UTP rather than STP.
Coaxial Cable
Coaxial cable, more commonly known as coax or cable TV cable, has specialized
insulators and shielding separating two conductors, allowing reliable, high-speed
data transmission over relatively long distances. Figure 3-3 illustrates a cross-section
of a typical coaxial cable. Coax comes in various thicknesses and has been histori-
cally used in Ethernet network architectures. In some cases, these network architec-
ture specifications include required characteristics of the (physical layer) coaxial
cable over which the (data-link layer) MAC layer protocol is transmitted.
Modern local area network implementations rarely use coaxial cable. However,
coaxial cable is still a key component of the cable television system. With the advent
of cable modems and the use of the cable television system as a mechanism to pro-
vide high-speed Internet connectivity to homes coaxial cable continues to play an
important role in data communication.
Fiber-Optic Cable
Fiber-optic cable is the current reliability and performance champion of the data
communication world. Although the most expensive media choice currently avail-
able, fiber-optic media delivers data transmission speeds measured in gigabytes (bil-
lions of characters) per second over distances often measured in miles.
Fiber-optic cable is also one of the most secure of all media, as it is relatively un-
tappable, transmitting only pulses of light, unlike all of the aforementioned media,
which transmit varying levels of electrical pulses. Because fiber optic is really a thin
Solid metal Plastic insulator Foil shield Braided Plastic or vinyl

inner core (usually white) shield/outer jacket
conductor
Figure 3-3 Coax Cable: Cross-Section

fiber of glass, it is immune to electro-magnetic interference, contributing to its high

bandwidth and data transmission capabilities. Another important thing to remember
is that fiber-optic cable requires more careful handling than copper media. Figure 3-4
illustrates a cross-section of a fiber-optic cable.
Light Transmission Modes Once a pulse of light enters the core of the fiber-optic cable,
it will behave differently, depending on the physical characteristics of the core and
cladding of the fiber-optic cable. In a Multimode or Multimode Step Index fiber-
optic cable, the rays of light will bounce off of the cladding at different angles and
continue down the core while others will be absorbed in the cladding. These multiple
rays at varying angles cause distortion and limit the overall transmission capabilities
of the fiber. This type of fiber-optic cable is capable of high bandwidth transmission
but usually over fairly short distances.
By gradually decreasing a characteristic of the core known as the refractive index
from the center to the outer edge, reflected rays are focused along the core more effi-
ciently yielding higher bandwidth over several kilometers. This type of fiber-optic
cable is known as Multimode Graded Index Fiber.
The third type of fiber-optic cable seeks to focus the rays of light even further so
that only a single wavelength can pass through at a time, in a fiber type known as
single mode. Without numerous reflections of rays at multiple angles, distortion is
eliminated and bandwidth is maximized. Single mode is the most expensive fiber-
optic cable, but can be used over the longest distances.
Core Thickness The thickness of fiber-optic cable’s core and cladding is measured in
microns (millionths of an inch). The three major core thicknesses are 50, 62, and 100
microns with their associated claddings being 125, 125, and 140 microns, respec-
tively. The increasing core thicknesses generally allow transmission over longer dis-
tances at a greater expense, however.
Light Source Wavelength The wavelength of the light that is pulsed onto the fiber-
optic cable is measured in nanometers (nm), with the optimal light transmitting
wavelengths coming in three distinct windows of 820 nm, 1,310 nm, and 1,500 nm;
820nm and 1,310nm are most often used for local and campus-wide networking
such as gigabit Ethernet, while 1,310nm and 1,500nm are used by carriers to deliver
high bandwidth fiber-based service over long distances. The higher-frequency light-
emitting sources carry a higher price tag.
Glass core Glass cladding Plastic or vinyl jacket
50 microns 125 microns

diameters
62 microns Multimode 125 microns

100 microns 140 microns
2-8 microns Singlemode
Note: A micron is a millionth of a meter
Figure 3-4 Fiber-Optic Cable: Cross-Section

Architectures
Gigabit Ethernet
Fast Ethernet
Ethernet
Distance
ATM
Media Type Also Called Bandwidth Limits Connectors Comments/Applications
Applied Problem
4-wire phone Quad RYGB 3 Kbps 200 feet RJ-11 jacks 4 insulated wired-
Solving station wire red,green, yellow, black.
Home phone wiring. Voice
Applications
Flat gray Flat satin, 14.4 Kbps 10-20 feet RJ-11 or Comes with 4,6,8
modular telephone RJ-45 plugs conductors. Used for
cable, silver short data cables using
satin modular (mod-tap)
adapters
Unshielded UTP 100 Mbps 100 feet RJ-45 5 Designated categories.

twisted pair Twists prevent
interference, increase
bandwidth. Voice grade
usually not suitable for
data
Shielded STP 16 Mbps 100 feet RJ-45 or IBM Shielding reduces

twisted pair data interference but
connectors complicates installation
Coax- thick Frozen 10 Mbps 500 feet AUI Original Ethernet cabling
yellow (attachment
garden hose unit interface)
Coax-thin RG-58, 10 Mbps 200 feet BNC connector Looks like cable TV cable.
thinnet, Easier to work with than
cheapernet thick coax.
Coax-thin RG-62 2.5 Mbps 200 feet BNC or IBM Similar to RG-58 (thinnet)
data connector but different electrical
characteristics make these
cables NOT
interchangeable
Fiber-optic Fiber several several SI or SMA 905 Difficult to install but

cable Glass Gbps kilometers or SMA 906 technology is improving.
High bandwidth, long
distance, virtually error
free, high security
Figure 3-5 LAN Media Technology Analysis
■ POINT-TO-POINT DATA TRANSMISSION TECHNOLOGIES
The most basic data communication technologies are those used to directly connect
two devices. These connections can be used to connect a computer to peripheral
devices such as modems, scanners, PDAs, and so on, or as the basis for a directly con-
nected computer-to-computer data connection. Operating at layer one of the OSI
Network Reference Model, these technologies provide a physical connection that can
be used to carry many higher-level protocols.
Serial Transmission Standards
Serial transmission is the basis of most data communication between computers.

There are several different serial communication standards available for use in mod-
ern computers, including RS-232, USB, and IEEE 1394 (Firewire).
Point-to-Point Data Transmission Technologies 79
RS-232 RS-232 is currently the most commonly used serial standard for modem
communication. Standardized by the Electronics Industries Association (EIA), RS-
232 is currently in its third release (RS-232-C). The prevalence of RS-232 in the PC
marketplace is so great that the term serial port has come to mean an RS-232 serial
connection. Although newer serial standards threaten to replace RS-232 as the pre-
dominate serial standard in the not too distant future, it still merits a close inspection.
RS-232 uses electrical signals to transmit the ones and zeroes of the digital datas-
tream. The RS-232 standard defines voltages of between +5 and +15 volts DC on a
given pin to represent a logical zero, otherwise known as a space, and voltages of
between 5 and 15 volts DC to represent logical ones, otherwise known as a mark.
RS-232 can be implemented on a multitude of data connectors. In the case of the
DB-25 connector illustrated in Figure 3-6, the presence or absence of an electrical
charge on each of these 25 pins has been designated as having a specific meaning in
data communications. These standard definitions are officially known as RS-232-C.
They were issued by the Electronics Industries Association (EIA), and are listed in
Figure 3-8.
Although all 25 pins are defined, in most cases, 10 or fewer of the pins are actu-
ally used in the majority of serial transmission applications. On some PCs, such as
1 2 3 4 5 6 7 8 9 10 11 12 13 1 2 3 4 5
14 15 16 17 18 19 20 21 22 23 24 25 6 7 8 9
Pin Pin Pin

Number Signal Designation Number Signal Designation Number Signal Designation
1 Protective ground 14 Secondary transmit data 1 Carrier detect
2 Transmit data 15 Transmit clock (DCE) 2 Receive data
3 Receive data 16 Secondary receive data 3 Transmit data
4 Request to send 17 Receiver clock 4 Data terminal ready
5 Clear to send 18 Receiver dibit clock 5 Protective ground
6 Data set ready 19 Secondary request to send 6 Data set ready
7 Signal ground 20 Data terminal ready 7 Request to send
8 Carrier detect 21 Signal quality detector 8 Clear to send
9 Positive DC test voltage 22 Ring indicator 9 Ring indicator
10 Negative DC test voltage 23 Data signal rate selector
11 unassigned 24 Transmit clock (DTE)
12 Secondary carrier detect 25 Busy
13 Secondary clear to send
Figure 3-6 RS-232 Serial Transmission Protocol as Defined for DB-25 and DB-9 Connectors
personal computers as well as many notebook and laptop computers, the serial port
has only 9 pins (DB-9 connector) and the RS-232 serial transmission protocol is sup-
ported as listed in Figure 3-6. Other RS-232 implementations rely on mini-DIN con-
nectors or phono-plug style connectors. Pictures of common RS-232 connectors are
shown in Figure 3-7.
PHYSICAL INTERFACES VS. TRANSMISSION PROTOCOLS

It is important to distinguish between those standards that describe the connectors or
physical interfaces that are used to connect appropriate cables to a computer’s phys-
ical ports, and the standards that describe the electrical characteristics, or transmis-
sion protocols used.
Practical Advice
and Information RS-232 Cables To connect devices via RS-232, a multi-wire cable must be used. The
cable has several small, insulated wires within an outer jacket. Each signal to be car-
ried, or RS-232 pin to be supported, requires its own individual inner wire.
The number of signals that must be transmitted across an RS-232 connection
between two devices will depend on the software used. The number of signals
required can vary from as few as 2 for one-way communication to 12 for a full-
duplex modem connection. Figure 3-8 summarizes the signals that are most com-
monly included in modem cables and designates which signals are assigned to
which pins on both DB-9 and DB-25 connectors. The signals are arranged in logical
pairs in order to increase understanding rather than in order of pin number.
DCE vs. DTE The original application of RS-232 was to connect computing devices
to modems. To make cabling this “standard” solution easier, two classifications of
RS-232 devices were created. The pin-outs for these classifications were established
so that the cable could be “straight through” in nature, where each pin is connected
to the same pin on the other end of the cable. Pin 1 will connect to pin 1, pin 2 to pin
2, and so on. The PC and the modem in our scenario are examples of these two clas-
sifications: DTE (data terminal equipment) and DCE (data communications equip-
ment), respectively. DCE can also expanded as data circuit terminating equipment.
Although this approach made sense for modem applications, the use of RS-232 has
expanded to include all sorts of electronic equipment ranging from modems to per-
sonal digital assistants (PDAs) to advanced television remote controls. It is therefore
critical to establish exactly which signals are associated with each pin on the connector.
It is impossible to rely on any standard mapping, as each device might have deviated
from the loosely enforced standard if they even use a “standard” connector.
DB-25 connector (female) DB-9 connector (female)
Figure 3-7 Common RS-232 Connectors

RS-232 DB- DB-

Signal 25 Pin 9 Pin Abbr. From To Explanation
Protective 1 5 PG A reference voltage used to protect circuit boards inside PC
Ground
Signal 7 5 SG A reference voltage used to determine proper signal voltage
Ground for ones and zeroes
Transmit 2 3 TXD DTE DCE Discrete voltages representing characters encoded as 1s and
Data 0s are transmitted on this pin to deliver the actual
data message
Receive 3 2 RXD DCE DTE Discrete voltages representing characters encoded as ones
Data and zeroes are received on this pin to receive the actual
data message
Request to 4 7 RTS DTE DCE Used in conjunction with CTS to perform modem-to-modem
Send flow control allowing modems to take turns transmitting to
each other.
Clear to 5 8 CTS DCE DTE Used in conjunction with RTS to perform modem-to-modem
Send flow control allowing modems to take turns transmitting to
each other
Data Set 6 6 DSR DCE DTE Used for initial handshaking between local modem and local
Ready PC to indicate local modem is functional
Data 20 4 DTR DTE DCE Used for initial handshaking between local modem and local
Terminal PC to indicate local PC is functional
Ready
Transmit 15 TC DTE DCE Clocking signal transmitted on this pin. Required for
Clock synchronous modems only.
Receive 17 RC DCE DTE Clocking signal received on this pin. Required for
Clock synchronous modems only.
Carrier 8 1 CD DCE DTE Indicates that the local modem has successfully contacted the
Detect remote modem and is ready to transmit data
Ring 22 9 RI DCE DTE Indicates to the local modem that a call is incoming and that
Indicator the modem should auto-answer the call.
Figure 3-8 Most Commonly Used RS-232 Signals
Serial/Parallel Conversion: UARTs Data travel via parallel transmission within a PC

over the PC’s main data bus. The data emerging from the serial port must be in serial
format. Therefore, somewhere inside the PC a parallel to serial conversion must be
taking place. A specialized computer chip known as a UART (Universal Asynchro-
nous Receiver Transmitter) acts as the interface between the parallel transmission of
the computer bus and the serial transmission of the serial port. UARTs differ in per-
formance capabilities based on the amount of on-chip buffer memory. The com-
monly used 16550 UART chip contains a 16 byte on-chip buffer memory for
improved serial/parallel conversion performance.
Physical
Standard Standards Interface
Name Body Connector Description
In Sharper Focus RS-422 EIA DB9 DB25 An electrical specification usually associated with
DB37 RS-449 (DB37 Connector). Each signal pin has its own
ground line (balanced) rather than sharing a common
ground. Up to 10Mbps over 1,200 meters. Use of DB25
or DB9 is also possible.
RS-423 EIA DB9 DB25 An electrical specification is usually associated with
DB37 RS-449 (DB37 Connector). Signal pins share a common
ground wire. (Unbalanced signaling) Up to 10 Mbps
over 1,200 meters. Use of DB25 or DB9 is also possible.
RS-449 EIA DB37 plus A physical/mechanical specification for a DB-37
DB9 (37-pin connector), plus an additional DB-9 if required.
Usually associated with either RS-422 or RS-423
electrical specifications.
RS-485 EIA DB9 DB25 Can be used in multipoint applications in which one
DB37 computer controls multiple (up to 64) devices. Often
used in computer-integrated manufacturing
operations or in telecommunications
management networks.
RS-530 EIA DB25 A physical/mechanical specification that works with
RS-422 or RS-423 over a DB-25 connector rather than a
DB-37 connector. Allows speeds of up to 2 Mbps.
V.35 ITU M-Block An international standard for serial transmission up to
48 Kbps defined for an M-block connector. Often used
on data communications equipment that must
interface to high-speed carrier services.
Figure 3-9 Other Traditional Serial Transmission Standards
OTHER TRADITIONAL SERIAL TRANSMISSION STANDARDS

RS-232 is officially limited to 20 Kbps (kilobits per second) for a maximum distance
of 50 feet. In reality, depending on the type of media used and the amount of external
interference present—RS-232 can be transmitted at higher speeds and/or over
greater distances. However, modern hardware typically supports speeds up to 115
Kbps using 16550 family UARTS.
To resolve the speed and distance limitations of RS-232, several other serial trans-
mission standards have been developed. Some of these standards also include the abil-
ity to connect more than two devices at the same time. A serial standard that can
connect three or more devices together concurrently is known as a multipoint serial
communications standard. A list of common serial standards is provided in Figure 3-9.
Universal Serial Bus (USB) Although RS-232 is historically the most widely imple-
mented serial standard, a new standard serial communication standard has virtually
replaced RS-232 in all but the most basic applications. This is because three basic archi-
tectural limitations in the RS-232 technology limit its application to new devices such
as scanners, digital cameras, video cameras, and personal digital assistants: RS-232 is
slow (only up to 115 Kbps), only supports one device per port, and requires significant
configuration to attach a device.
The universal serial bus (USB) is a high-speed, multipoint serial communications
technology developed by the USB Implementer’s Forum to resolve RS-232’s shortcom-
ings. There are two versions of USB currently available—the original USB 1.1 specifica-
tion and a newer, higher-speed USB 2.0 specification. USB 2.0 is backward compatible
with USB 1.1. However, to gain the higher speeds offered by USB 2.0, devices must be
attached either directly to the computer or to a USB 2.0 compatible hub.
USB version 1.1 operates at either 1.5 Mbps or 12 Mbps. That equates to more
than 1,000 times the maximum speed of RS-232. USB 2.0 can operate at even higher
speeds, up to 480 Mbps. This increased speed is significantly important when sup-
porting such high data devices as scanners and digital cameras. As covered in chap-
ter 4, it is even possible to connect to a local area network (LAN) via a USB network
interface card (NIC).
Both USB 1.1 and USB 2.0 use electrical signals to transmit the ones and zeroes of
the digital datastream. Unlike RS-232, which is ground referenced (uses voltages of
+5Vdc and 0Vdc for ones and zeroes), USB uses a differential electrical signal. A digital
1 is represented when D+ signal line voltage is 200 mv greater than the D– signal volt-
age. Similarly, a digital 0 is represented by the D+ signal voltage being 200 mV less than
the D– signal voltage. Differential signaling offers more noise resistance than ground
referenced signaling while greatly reducing the potential impact of ground loops.
When connected to a computer, low-speed USB devices such as joysticks and PDAs
create virtual serial ports. These ports are typically assigned as if they were physical RS-
232 com ports (COM2, COM3, etc.) Since each of these ports can only use up to 115 Kbps
of bandwidth, they can easily be multiplexed over the same physical USB connection
with no loss of performance. However for high-speed devices such as scanners and
video cameras, it is better to connect them to one of the computer’s USB ports.
USB can support up to 126 devices on each port. Some devices come with two
USB ports so that the user can simply “daisy-chain” the devices together. Although
this solution is functional for a couple of devices, a better solution for larger imple-
mentations is to use a USB hub. As illustrated in Figure 3-10, a USB hub is a device
that connects directly to the computer’s USB port and offers multiple ports to con-
nect other USB devices.
computer with
USB port
4 port 4 port
USB hub USB hub
scanner joystick
digital
camera PDA
MP3
player
Figure 3-10 USB Hub Implementations

To reduce the configuration hassles traditionally associated with serial connectivity

using RS-232, USB is designed to be plug and play. Devices are automatically given an
address, assigned a speed, and identified by the computer’s operating system. As long
as the operating system supports USB plug and play all that is required to connect a
device is to plug it into either the computer or to a USB hub connected to the computer.
To make the process of attaching devices even easier, the USB ports and cables
also include the power to the device, eliminating the need to have a separate power
supply for each peripheral. When using powered USB devices it is important that the
power consumption of the attached devices does not exceed that of the computer port
or USB hub to which the devices are attached. While the USB specification requires
500 milliamps of power, in an effort to save battery life some notebook computer ven-
dors have reduced the USB power output to lower levels. In these cases, some combi-
nations of devices may not work without adding an externally powered USB hub.
There are two connector types associated with USB. The USB “A” connector is
used to connect to computers and the USB “B” connector is used to connect to the
devices. The ability to disconnect the USB cable at the device end of the connection
makes it significantly easier to change devices compared to having to reach behind
the computer to disconnect the cables. Figure 3-11 shows the USB A and B connectors.
IEEE-1394/Firewire Another high-speed serial standard is IEEE 1394. Originally

developed by Texas Instruments and implemented by Apple Computer Inc. as the
proprietary Firewire, the interface was standardized by the IEEE in 1995. Sony has
trademarked the name i.Link for their implementation of IEEE-1394, and that
moniker seems to be gaining popularity, at least among Japanese consumer electron-
ics manufacturers.
Similar to USB, IEEE-1394 is a multipoint serial bus–based solution. Devices can
be added or removed from the live bus. Devices can be daisy chained or connected to
an IEEE-1394 hub for connectivity between more than two devices. Although similar
to USB in many ways, IEEE-1394 is not intended to directly compete with USB but
rather to complement USB. USB offers a low- to medium-speed serial solution while
IEEE-1394 offers a high-speed solution. By having two separate serial standards,
each can be optimized for specific applications.
IEEE-1394 uses electrical signals to transmit the ones and zeroes of the digital
datastream. Like USB, IEEE-3194 uses differential signaling to provide better perfor-
mance and noise abatement over longer distances. Unlike USB IEEE-1394b is also
capable of operating over fiber-optic media, offering higher speeds and greatly
increased distances.
USB-A (on left) and USB-B (on right) connectors
Figure 3-11 USB Connectors

IEEE-1394 is a high-speed serial solution ranging in speed from 400 Mbps (origi-
nal specification) to 800 Mbps (1394b). One of the key differentiators between IEEE-
1394 and other serial bus standards is that, in addition to standard asynchronous
communication, IEEE-1394 includes support for isochronous communication. Isochro-
nous communication guarantees data delivery at a constant, predetermined rate. This
allows IEEE-1394 to be used in time critical multimedia solutions. The constant data
delivery rate reduces the need to buffer data, thereby greatly reducing the cost of
implementing the technology compared to a traditional asynchronous solution.
Taking advantage of its isochronous capabilities, the main applications of IEEE-
1394 at the current time are in the transmission of digital data between consumer
electronics devices. IEEE-1394 has been adopted by the digital VCR Manufacturers
Association as its standard interconnection technology. The ongoing conversion of
television from an analog format to digital HDTV technologies will increase the mar-
ketplace for IEEE-1394 technologies in the consumer electronics marketplace.
IEEE-1394 ports are also being introduced in multimedia-capable computers. In
addition to allowing the computer to connect to digital camcorders and VCRs, IEEE-
1394 ports will allow data network solutions to be deployed. To support the use of
IEEE-1394 on computer platforms, the IEEE-1394 Trade Association released an
updated version of the standard known as IEEE-1394b in 2002. The IEEE-1394b stan-
dard increases speeds to 800 Mbps at 100 meters and reduces latency by streamlining
the protocol. Also known as Firewire 800, IEEE-1394b is designed to be backward
compatible with the original IEEE-1394 standard.
IEEE 1394 uses two different connectors: a four-wire connector that supports
full-duplex data transmission and a six-wire connector that also provides power to a
connected device. Figure 3-12 shows these two interface connectors.
Parallel Transmission
Parallel transmission is primarily limited to transmission of data within a computer,

although the release of the IBM personal computer included an 8-bit parallel inter-
face designed to connect the computer to a printer. Common physical interfaces
associated with this parallel transmission technology are the DB-25 connector and
the Centronics connector. The Centronics connector is a 36-pin parallel interface. In
addition to the physical plug and socket, the Centronics parallel standard defines
electrical signaling for parallel transmission and is a de-facto standard. DB-25 and
Centronics parallel physical interfaces are illustrated in Figure 3-13.
IEEE-1394 Connectors
Figure 3-12 IEEE-1394 Connectors

DB-25 (male) parallel interface Centronics parallel interface
Figure 3-13 Parallel Transmission Physical Interfaces
Initially intended for connecting a printer, the parallel port on personal comput-
ers later saw duty as a means of connecting medium-speed peripheral devices such
as scanners, tape drives, and video cameras due to its relatively high speed as com-
pared to RS-232. However, with the proliferation of USB and IEEE 1394, use of the
Centronics parallel port has significantly trailed off to the point that some notebook
computer vendors are no longer including parallel ports on their products.
Wireless Technologies
One of the largest hassles with the explosion of modern computing devices is the
series of cables required to connect them. This is even more of a problem when trav-
eling. If you have a notebook computer, a PDA, a cell phone, and an MP3 player, you
could potentially need three different data cables to connect them, in addition to the
requisite power supplies and cords. Two technologies have been developed as wire-
less cable replacements: Infrared (IRDA) and radio (Bluetooth).
Infrared (IrDA) The Infrared Data Association (IrDA) is an international standards

body responsible for creating standards that apply to the use of infrared (Ir) light to
provide wireless connectivity for devices and/or to locations that would typically
require more traditional wire-based solutions. IrDA signals are point-to-point nature,
have a narrow angle (30 degree cone), are limited to around meter, and have a
throughput of 9,600 bps to 4 Mbps. IrDA has proven to be a popular technology, with
compliant ports currently available in an array of devices, including embedded
devices, pagers, phones, modems, cameras, watches, computers (PCs) and laptops,
PDAs, printers, and other computer peripherals.
There are two widespread IrDA versions: IrDA 1.0 and 1.1. Designed as a direct
RS-232 serial port replacement, IrDA 1.0 (also known as SIR) offers data speeds of 9.6,
19.2, 38.4, 57.6, and 115.2 Kbps. IrDA 1.1 (FIR) increases the list of available speeds to
include 576 Kbps, as well as 1.152 and 4.0 Mbps. A third version of the standard (FIR)
that supports data speeds up to 16 Mbps was ratified but has not seen widespread
use. Regardless of version, IrDA ports are addressed as RS-232 serial ports.
Despite widespread acceptance, infrared has some significant limitations to its
use. In general, IrDA is a point-to-point protocol; only two devices can communicate
at any given time. If a session is live between two devices and a third device attempts
to initiate a session, the initial session will be dropped. There is a derivative of IrDA
that supports multi-point communication (IrLAN), but it has not achieved wide-
spread adoption and probably will not due to the evolution of Bluetooth technologies.
Internet Access Technologies 87
Another key limitation of IrDA is that there must be a direct line of sight between
the two communicating devices. If either is moved, the communication session will
be dropped. This limitation effective prevents IrDA from being used in many appli-
cations such as connecting a PDA to the Internet through a cell phone in a belt hol-
ster. Using IrDA, the phone must be out on a table right in front of the PDA. Between
these limitations and the limited speed available, IrDA is likely to be replaced by
Bluetooth as that standard matures.
Bluetooth Bluetooth is the name given to an emerging wireless radio frequency (RF)
communication standard. Originally developed by Erickson, Bluetooth is currently
backed by Bluetooth Special Interest Group (SIG), a collection of Bluetooth promot-
ers that includes 3COM, IBM, Intel, Microsoft, and Nokia, among others, with the
purpose of developing and promoting the use of the Bluetooth technology.
Bluetooth uses radio frequency communication in the unlicensed 2.4-2.4835 GHz
band using frequency hop spread spectrum (FHSS). As its name suggests, FHSS hops
from one frequency to another throughout the allowable frequency range. The pat-
tern of frequency hopping must be known by the wireless receiver so that the mes-
sage can be reconstructed correctly. A given wireless transceiver’s signal is on a given
frequency for less than 1 second. By constantly changing frequency, the transmission
tends to be less affected by interference, an especially desirable characteristic for
mobile computing applications.
Bluetooth offers data speeds of up to 1 Mbps up to 10 meters. Unlike IrDA, Blue-
tooth supports a LAN-like mode where multiple devices can interact with each other.
Due to the limited size of the Bluetooth coverage area, the terms piconet and personal
area network (PAN) have been coined to describe a Bluetooth coverage zone.
The key limitations of Bluetooth are security and interference with wireless
LANs. The fact that Bluetooth doesn’t require line-of-sight to communicate cuts both
ways: while it makes it possible to use your PDA with your cellular phone that is
stashed in your briefcase, it also makes it possible for someone else with a device hid-
den from view to attempt to gain illicit access to your computing devices. Bluetooth
attempts to mitigate this risk through the use of strong authentication and encryp-
tion technologies. Because Bluetooth uses the same frequency range as IEEE 802.11b
wireless networks, simultaneous use of both technologies greatly affects perfor-
mance. This concept is covered further in chapter 4.
Devices hitting the market that include Bluetooth include cellular phones, PDAs,
headphones, and mobile gaming platforms. Although the technology is too new to
guarantee widespread success, it shows great promise as a way to eliminate the
spaghetti bowl of cords that beleaguers mobile workers.
■ INTERNET ACCESS TECHNOLOGIES
After connecting to peripheral devices, most people want to connect their computer
to the Internet. While a technical discussion of the Internet is beyond the scope of this
chapter, suffice it to say that it is a large, international network that offers a wide-
ranging selection of data services to any connected computer including e-mail,
instant messaging, Web pages, and streaming media. More detail on the technical
“how” of the Internet will be provided in chapters 7 and 8.
The process of connecting to the Internet is conceptually straightforward: All
you have to do is create a link between your computer and a device connected to the
Internet. The traffic from the local computer that needs to go to the Internet will be
sent across this link. The problem is the distance between the local computer and the
Internet-connected device.
Although the point-to-point data communications technologies previously dis-
cussed provide a means to connect a computer to a local device or even another com-
puter located in close physical proximity, they are not capable of dealing with the
distance between the local computer and the remote Internet connection device. A
whole new class of data communication technologies is required to meet the distance
and speed requirements for Internet access. The Internet access architecture is illus-
trated in Figure 3-14.
Internet Service Providers
In most cases, an individual wishing to connect to the Internet will initiate a link to a
company whose business is extending the Internet to individuals. These companies,
known as Internet service providers (ISPs), take a high-speed connection to the
Internet and—using a collection of Internet access technologies—offer Internet con-
nectivity to individual users.
ISPs range in size from a local provider that focuses on meeting the needs of a
geographically limited area to national and international providers such as America
Online (AOL), which provide service across large geographic areas. In addition to
the area they service, ISPs vary widely in the access technologies offered, data speeds
provided, and pricing methods.
SELECTING AN ISP
Regardless of size, ISPs all provide access to the same Internet. When selecting an ISP
it is important to look at four key criteria:
Managerial
Perspective • Service hosting—Do you need the ISP to host your Web pages, e-mail
addresses, and so on?
• Performance—What types of access technologies does the ISP support? What
data rates are offered? What is the confirmed information rate (CIR—the
Corporate Office
Internet
server
Consumer last
mile T-1, last
mile access
computer frame relay, access device
access ATM, server
server Metro Ethernet (www.slashdot.org)
access phone
device line router
USB/ CSU/DSU
modem LAN
fire wire modem cable
bank, ethernet
cable/DSL line
DSLAM,
CMTS
network protocols
Figure 3-14 Internet Access Architecture

amount of throughput the ISP guarantees, which is usually significantly

slower than the base speed)?
• Cost—What is the monthly cost of the service? What is the cost per megabyte
per second? Are there service caps after which surcharges apply?
• Reliability—Does the ISP have a generator on site to maintain connectivity in
the event of a power failure? Does the ISP have multiple links to the Internet
so that if one fails you still have access? Does the ISP offer multiple links to
your location to protect from the failure of a single link?
Dial-Up Modems
The first and most basic method for connecting a home computer to the Internet is to
use a dial-up modem. A dial-up modem creates a circuit-switched connection across
the public switched telephone system (PSTN) to another modem that is connected to
the Internet.
Architecture A switched or dial-up line is the type of phone line typically installed in
a home or place of business. To place a call, you pick up the receiver or handset, wait
for a dial tone, and dial the number of the location you wish to call. This ordinary type
of phone service is sometimes called POTS or Plain Old Telephone Service. More
formally, the phone network is referred to as the Public Switched Telephone Net-
work (PSTN). Figure 3-15 illustrates the major components of the PSTN.
As introduced in chapter 1 and detailed in chapter 5, the phone set is connected
to a large switch in a telephone company building called a central office or CO via a
local loop. When a call is dialed, the telephone switch connects the phone set to the
destination set by finding an available circuit or path.
Technology Calls placed over dial-up lines through CO switches that have connec-
tions built from available circuits are called circuit-switched connections. In order to
interface transparently to the PSTN, modems must be able to dial and answer phone
calls to and from other modems.
The next important characteristic related to transmitting data over a dial-up
phone line has to do with how the data are represented on that phone line. First, it is
Local PC C.O. Remote PC

Local loop
Modem Modem
Local loop
Inter-exchange
circuit
P.O.P.
P.O.P. C.O.
Figure 3-15 PSTN Architecture

important to realize that today’s dial-up phone network was originally designed to
carry voice conversations efficiently and with reasonable sound quality. This “effi-
ciency of design with reasonable sound quality” meant reproducing a range of the
frequencies of human speech and hearing just wide enough to produce reasonable
sound quality. That range of frequencies, or bandwidth, is 3,100 Hz (From 300 Hz to
3,400 Hz), and is the standard bandwidth of today’s voice-grade dial up circuits
(phone lines). Hz is the abbreviation for hertz. One hertz is one cycle per second. The
higher the number of hertz or cycles per second, the higher the frequency. Frequency,
wavelength, hertz, and cycles per second were introduced in chapter 2.
This 3,100 Hz is all the bandwidth with which the modem operating over a dial-
up circuit has to work. Because the dial-up phone network was designed to be able to
mimic the constantly varying tones or frequencies that characterize human speech,
only these continuous, wave-like tones or frequencies can travel over the dial-up
phone network in this limited bandwidth.
The challenge for the modem is to represent the discrete, digitized ones and
zeroes from the input (PC) side of the modem in a continuous or analog form within a
limited bandwidth so that the data may be transmitted over the dial-up network. Fig-
ure 3-16 summarizes the results of I-P-O analysis involving modems and the PSTN.
The modem’s job is to convert digital data into analog data for transmission over
the dial-up phone network and to convert analog data received from the dial-up net-
work into digital data for the terminal or PC. As previously discussed in chapter 2,
the proper names for these processes are modulation and demodulation. In fact, the
word modem is actually a contraction for Modulator/demodulator.
Most local loops that are used for connection to the PSTN to supply switched,
dial-up phone service are physically described as two-wire circuits. Since one of these
two wires serves as a ground wire for the circuit, that leaves only one wire between
the two ends of the circuit for data signaling. Dial-up or switched two-wire circuits
generate a dial tone.
Given that two-wire dial-up circuits only have one wire for data signaling, only
one modem could be transmitting at a time while the other modem could only be
receiving data. This one direction at a time transmission is known as half-duplex.
Modems that interfaced to a dial-up circuit had to support this half-duplex trans-
mission method. What this meant was that once the two modems completed initial
ISP
Home
Local PC Router
1000001 1000001
PSTN
Internet
Phone network
modem modem
Input Processing Output - Processing Output

Digital data Transform digital data Input Transform analog data Digital data
input to analog data Analog input to digital data
output (modulation) output (demodulation)
Figure 3-16 I-P-O Analysis: Modems and the PSTN

handshaking, one modem would agree to transmit while the other received. In order
for the modems to reverse roles, the initially transmitting DTE (terminal or com-
puter) drops its RTS (request to send) RS-232 Pin # 4 and the transmitting DCE
(modem) drops its CTS (clear to send) RS-232 Pin # 5, and perhaps its carrier wave.
Next, the initially receiving DTE must raise RTS, the initially receiving DCE (modem)
must generate a carrier wave and raise CTS, and the role reversal is complete.
This role reversal is known as turnaround time and can take two-tenths of a sec-
ond or longer. This might not seem like a very long time, but if this role-reversal
needed to be done several thousand times over a long-distance circuit, charged by
usage time, it might have a large dollar impact.
Full-duplex transmission supports simultaneous data signaling in both directions.
Full-duplex transmission might seem to be impossible on two-wire circuits. Until the
advent of the V.32 9,600 bps full-duplex modem, the only way to get full-duplex trans-
mission was to lease a four-wire circuit. Two wires (signal and ground) were for trans-
mitting data and two wires (signal and ground) were for receiving data. There was no
“role reversal” necessary and therefore, no modem turnaround time delays. Modems
manufactured to the CCITT’s V.32 standard (and the later V.34 standard) can transmit
in full-duplex mode, thereby receiving and transmitting simultaneously over dial-up
two-wire circuits using a sophisticated echo cancellation technique.
Echo cancellation takes advantage of sophisticated technology known as digital
signal processors (DSP) that are included in modems that offer echo cancellation. By
first testing the echo characteristics of a given phone line at modem initialization
time, these DSPs are able to actually distinguish the echoed transmission of the local
modem from the intended transmission of the remote modem. By subtracting or can-
celing the echoed local transmission from the total data signal received, only the
intended transmission from the remote modem remains to be processed by the
modem and passed on to the local PC.
Standards In order for modems manufactured by different vendors to interoperate

successfully, they must adhere to common operational standards. These standards
must define methods of modulation, data compression, error correction, and auto-
dialing, and must provide backward compatibility with older standards. Figure 3-17
summarizes significant dial-up modem standards of the last 20 years.
The V series standards featured in Figure 3-17 are officially sanctioned by ITU-T
(International Telecommunications Union—Telephony Sector). The Bell standards
listed in Figure 3-17 are pre-divestiture standards from the time when AT&T dictated
modem specifications in the United States and compliance with international stan-
dards was not an option. The suffix “bis” refers to the second standard issued by a
given standard committee, while the suffix “ter” refers to the third standard issued
by that same committee.
NON-STANDARD STANDARDS
In an effort to gain market share during the time when standards-making organiza-
tions are deliberating new standards, vendors often introduce proprietary versions
of pending standards. Before the standardization of V.90 there were two proprietary
56 Kbps modem technologies: X2 from 3COM/US Robotics and K56flex from Lucent
Practical Advice Technologies.
and Information One very important point to keep in mind when purchasing pre-standard data
communications equipment is the ability of the vendor to upgrade that equipment to
Modem Transmission Data Error Modulation

Standard Rate Baud Rate Compression Correction Method
V.90 56 Kbps 3,200, V.42bis/MNP5 V.42/MNP4 Digital downlink &
down, 28.8 3,000, 9QAM &TCM uplink
Kbps up 2,400,
2,743,
2,800,
3,429, baud
V.34 28.8 Kbps 3,200, V.42bis/MNP5 V.42/MNP4 9QAM &TCM
33.6 Kbps 3,000,
(optional) 2,400,
2,743,
2,800,
3,429, baud
V.32 ter 19.2 Kbps 2,400 baud V.42bis/MNP5 V.42/MNP4 8QAM & TCM
V.32 bis 14.4 Kbps 2,400 baud V.42bis/MNP5 V.42/MNP4 6QAM & TCM
V.32 9.6 Kbps 2,400 baud V.42bis/MNP5 V.42/MNP4 4QAM & TCM
V.22 bis 2,400 bps 600 baud V.42bis/MNP5 V.42/MNP4 4QAM & TCM
Bell 212A 1,200 bps 600 baud 4PSK
Bell 103 300 bps 300 baud FSK
Figure 3-17 Modem Standards
meet the specifications of the official standard once it is issued. In some cases, soft-
ware upgrades are possible, while in other cases, hardware upgrades or chip replace-
ment is required. In some cases, these upgrades may be free and easily accomplished
via the Internet, while in other cases, the upgrade may involve returning the equip-
ment to the factory, involving upgrade fees of several hundreds of dollars. Be sure to
understand all of the details regarding standards compliance upgrades before ever
purchasing pre-standards data communications equipment.
V.34 V.34 offers a transmission rate of up to 33.6 Kbps (33,600 bps) over the stan-
dard analog local loop. The highest required speed for a V.34 compliant modem is
28.8 Kbps. The 33.6 Kbps speed is available for modems that make use of an optional
baud rate. As shown in Figure 3-17, the higher speed is achieved by using a lower
baud rate with more bits per baud than the required 28.8 Kbps speed.
It is important to note that although modulation standards have changed in
order to produce higher transmission rates, the associated data compression and
error correction standards have remained constant. The V.34 modem standard insti-
tuted a variety of technical innovations in order to achieve this transmission rate
over dial-up lines of variable quality. The overall effect of these technical innova-
tions is that the V.34 modem is better able than any previous modem standard to
easily and dynamically adjust to variable line conditions in order to optimize trans-
mission rate.
V.34 TECHNICAL INNOVATIONS

Figure 3-18 summarizes the technical innovations introduced with the V.34 modem
standard and their associated importance or implication.
In Sharper Focus Although all of the technical innovations listed in Figure 3-18 are considered part
of the V.34 standard, it is not safe to assume that they are all included in any V.34
modem. In addition, several other optional features may or may not be supported on
any given V.34 modem:
• Support of leased lines as well as dial-up lines

• Inclusion of four-wire as well as two-wire physical interfaces
• Password protection and callback security
V.34 Technical Innovation Importance/Implication

Multiple baud rates/ The V.34 standard specifies three required baud rates (3,200, 3,000, 2,400) and three
Carrier frequencies optional baud rates (3,429, 2,800, 2,743). Multiple potential baud rates increase the V.34
modem’s ability to adapt to variable line conditions. Each baud rate supports two
different carrier frequencies in order to optimize transmission speed on dial-up lines of
variable quality.
Baud rates greater The V.34 standard is the first to attempt baud rates of greater than 2,400. The 28.8Kbps
than 2,400 transmission rate is achieved by interpreting 9 data bits per baud at a baud rate of 3,200.
The 33.6 Kbps transmission rate is achieved by interpreting 12 bits per baud at a baud
rate of 2,800.
Auxiliary management An auxiliary or side channel, separate from the data channel, is available for
channel transmission of management or configuration data. This channel would be particularly
important if the V.34 modem were attached to a router or similar internetworking
device that might require monitoring or management without disrupting the main
data traffic.
Asymmetric transmit/ Some applications such as database queries/responses, or Web requests/downloads,
receive speeds may require wide bandwidth in one direction only. V.34 specifies a method for
allocating the data with a larger bandwidth in one direction than the other.
Adaptive line probing Adaptive line probing tests the characteristics of the transmission line not just at call
initiation time, but throughout the transmission. Changes in baud rate, carrier
frequency, constellation size and shape and other parameters can be changed in order to
optimize transmission rate.
Precoding and Reduces high-frequency line noise and increases immunity to interference on analog to
nonlinear encoding digital conversion by plotting constellation points in areas with less interference.
Fallback/Fallforward Although many modem standards support fallback, V.34 supports fall-forward features
that allow increases in transmission rates as line conditions improve.
Trellis coded modulation A forward error correction methodology that helps support higher baud rates on dirty
phone lines by predicting the location of a given constellation point. Can add significant
processing overhead.
V.8 Training specification Also known as fast-training, this specification allows two V.34 modems that support V.8
training to set-up and initialize a call faster than modems that do not support V.8.
Figure 3-18 V.34 Technical Innovations

• Ability to connect to fax machines via the V.17 FAX standard

• Ability to auto-dial via the V.25bis autodial standard
• Auto-dial backup for failed leased lines or lost carrier
• Auto restoral to repaired leased lines
This lack of assurance as to the exact features supported on any given V.34 can
lead to interoperability problems among V.34 modems. The best solution to this
dilemma is to purchase identical modems from a single manufacturer. If this is not
practical, then the time invested to carefully investigate which features of the V.34
standard are implemented in a given modem would be a wise investment indeed.
V.90 According to Shannon’s Law the highest V.34 transmission rate of 33.6 Kbps
should be the maximum transmission rate that a voice-grade telephone line can sup-
port. However, Figure 3-17 claims that modems that adhere to the V.90 standard are
capable of transferring data in at least one direction at speeds up to 56 Kbps per sec-
ond. How can this be? Were telecommunications engineers able to change the laws of
physics?
The answer is obviously no. Shannon’s law is still intact for data transmission
across the PSTN. The telecommunications engineers instead changed the paradigm
in order to alleviate one of the largest components of noise in a traditional modem
implementation. The resulting standard, known as V.90, is a hybrid analog/digital
standard that offers transmission rates of almost double V.34 levels.
A preview of chapter 5 shows that the majority of the telephone system is cur-
rently digital in nature with the local loops from the central offices to the end users
being the only remaining analog connections. An analog telephone signal is con-
verted into a digital signal at the Central Office (CO) for transmission across this dig-
ital telephone system then converted back into an analog signal at the destination CO
for transmission to the destination telephone across the destination local loop.
The limiting factor in exceeding V.34 speeds is the quantization noise associated
with the conversion of the analog signal to a digital signal for transmission across the
digital telephone system. In order to exceed V.34 speeds this analog to digital con-
version and its associated quantization noise must be eliminated. Fortunately the
digital nature of the modern PSTN makes this possible.
QUANTIZATION NOISE
Conversion of a signal from an analog format to a digital format imparts “noise” into
the signal. This can be readily explained by considering the nature of both analog
In Sharper Focus and digital signals. Analog signals can consist of any possible level. Digital signals
can consist only of fixed levels.
As shown in Figure 3-19, when an analog signal is sampled and converted into a
digital signal, a certain level of detail is lost. When the digital signal is converted
back into an analog signal at the destination end of the call, the resulting analog
waveform is not exactly equal to the source waveform. This error is known as quan-
tization noise or quantization error.
While quantization noise is not perceivable by humans on normal telephone
calls, it affects the ability of modems to use the entire bandwidth of the call, thus lim-
iting the potential transmission rate. It is important to note that quantization noise is
created by the analog to digital conversion process, although it does not manifest
itself until subsequent digital to analog conversion.
Original Analog Waveform
digital data stream
..0010011000100011..
Digital Sampling Recreated Analog Waveform
quantization noise
NOTE: graph exaggerated
for illustration purposes
Figure 3-19 Quantization Noise
The V.90 modem standard makes use of the digital PSTN to overcome quantiza-
tion noise by replacing the analog modem on one end of the connection with a digi-
tal server. This server connects directly to its Central Office (CO) via a digital
connection such as an ISDN PRI or T-1 line. By using a direct digital connection, the
server avoids any analog to digital conversions for outgoing transmission, thereby
eliminating quantization noise. At the other end of the connection the CO converts
the digital signal into a clean analog signal and transmits it to the destination
modem. This process is illustrated in Figure 3-20.
By using this technique, a theoretical transmission rate of 56 Kbps can be achieved
from the server to the modem. Note that this is a theoretical rate. An FCC limit on the
amount of power allowed on the local loop reduces this theoretical limit to around 53
Kbps. Physical limitations in the local loop—including the distance from the modem
to the CO, the number of connections or splices in the local loop, and the overall qual-
ity of the local loop—can limit the actual transmission rates further. In the authors’
experience it is far more common to achieve transmission rates of around 44 Kbps,
although connections at rates in excess of 50 Kbps have been noted.
C.O. C.O.
digital/analog NO
digital
digital analog/digital
analog
conversion -or-
digital/analog
analog/digital ISDN/T-1
V.90 modem PSTN
local loop V.90 server
conversion
conversion
Figure 3-20 V.90 Implementation
A close examination of Figure 3-20 shows that the increased speed offered by
V.90 is limited to the half of the connection originating at the server and terminating
at the modem. The other half of the connection, originating at the modem and termi-
nating at the server, must undergo an analog to digital conversion and therefore
incurs quantization error, limiting speeds to V.34 levels. Because the speeds are dif-
ferent in each direction, V.90 is said to be an asymmetrical transmission technology.
Fortunately, the asymmetrical characteristic of V.90 matches the asymmetrical
nature of the most common modem application: accessing the Internet. In a normal
Internet session, a user sends a fairly short request (such as the address of a Web
page) to a server. The server responds with a much larger collection of data (such the
text and graphics that make up the Web page) to the user; making V.90 a natural fit
for Internet access.
As with all modern modem standards, V.90 compliant modems are backward
compatible with earlier standards. When a connection is initially made, the modem
will attempt to connect at the highest possible V.90 speed. It will then drop back to
slower V.90 levels, then resort to V.34 levels, V.32 levels, and so on, until a carrier
wave can be established. Because of this automatic rate fallback strategy, it is com-
mon for first-time users to comment on the long time required for a V.90 modem to
establish a carrier and begin communicating.
Data Compression and Error Correction Standards By operating more efficiently,

modems are able to offer more transmitted data in a given amount of time. Although
several factors can lead to increased modem efficiency, data compression can have
the most significant impact in the amount of data actually delivered by a given
modem in a given amount of time.
Data Compression As introduced in chapter 2, data compression replaces large
strings of repeating character patterns with a special code that represents the pattern.
The code is then sent to the other modem. From that point forward, the sending
modem sends the code instead of the original pattern. As the code is significantly
smaller than the pattern it represents, the amount of data sent between the two
modems is reduced, up to 400 percent under optimal conditions. The process of data
reduction is similar in concept to having more than one bit per baud, as covered in
the previous chapter.
By using data compression, a 28.8Kbps (V.34) modem could optimally transfer
115.2Kbps (28.8 Kbps × compression factor of 4:1) of data across a dial-up phone line.
In this scenario, the 115,200 bps is known as throughput, while the transmission rate
remains at or below the V.34 maximum of 28.8Kbps.
V.42bis and MNP 5 The two predominant data compression standards are V.42 bis
and MNP5. Most modems try to negotiate with each other to implement V.42bis data
compression at initialization. The less-efficient MNP5 protocol is commonly used as
a second option.
V.42bis uses a data compression algorithm known as the Lempel Ziv algorithm.
Ideally, it can compress files, and thereby increase throughput by a 4:1 ratio. Proprietary
improvements to this algorithm by modem manufacturers can be achieved in two ways:
• Increase the amount of memory dedicated to the library, also known as the
dictionary (1.5KB standard, some modems use up to 6KB).
• Increase the size of the pattern of characters, also known as string size, which
can be stored in the dictionary (32 bytes standard, some modems support
strings up to 256 bytes).
Proprietary improvements to standards such as V.42bis are only effective when

both modems involved in a transmission are identical. It should also be noted that
most independent modem testing suggests that compression ratios in the range of
2.5:1 are most likely, despite higher optimal claims.
MNP Class 5 yields data compression ratios in the range of between 1.3:1 and
2:1. MNP5 uses two data compression algorithms:
• Huffman encoding is a special character encoding scheme that re-encodes

ASCII characters. Frequently used characters such as “a”, “e” and “s”, are
encoded with only 4 bits, while rarely occurring characters such as “x” or “z”
are encoded using as many as 11 bits. Overall, the effect of Huffman encoding
in that more characters are transmitted using fewer bits.
• Run-length encoding exams a datastream in search of repeating characters.
When any character repeats more than three times, the run-length encoding
algorithm replaces the entire string of repeated characters with only three
repetitions of the character followed by a count field indicating how many
times the character is actually repeated. For example, a data string containing
10 consecutive repetitions of the same character would be replaced by 3 repe-
titions of that character followed by a 1-byte count character. This would
reduce the string in question from 10 bytes to 4 bytes, for a savings of 60 per-
cent. Repeated characters can include nonprinting characters such as spaces,
carriage returns, and line feeds.
Improving the reliability of the data transmission link between two modems will
ultimately make that data transmission faster and more efficient. Fewer retransmis-
sions due to data errors will reduce the overhead associated with correcting errors,
thus reducing the time necessary to transmit a given message. This will minimize the
cost of the data transmission.
Error Prevention Data transmission errors occur when received data is misinter-
preted due to noise or interference on the phone lines over which the data message
traveled. Errors can be prevented by
• Reducing the amount of noise or interference on a given transmission line

• Employing modulation techniques that are able to adapt to and overcome
noisy lines
Line Conditioning, Repeaters, and Amplifiers Line conditioning is a value-added

service available for analog leased lines from the phone company. Various levels of
conditioning are available at prices that increase proportionally to the level of condi-
tioning, or noise reduction, requested. Conditioning represents a promise from the
phone company in terms of the noise levels or interference present on a given analog
leased line. In order to deliver on this promise, the phone company might have to
install additional equipment.
A repeater is often used by a phone company to assure signal quality over the
entire length of a circuit. As a signal travels through a medium such as copper wire,
it loses some of its strength due to the resistance of the wire. This loss of signal
strength or volume is known as attenuation.
A repeater on an analog circuit, sometimes called an amplifier, strengthens and
repeats the signal. Unfortunately, on an analog circuit the amplifier cannot distin-
guish between the voice or data signal and the background noise. As a result, the
amplifier strengthens and repeats both signal and the background noise.
Repeaters on digital circuits are able to distinguish the digital signals from the
background noise. Therefore, digital repeaters can retransmit a digital signal free of
noise, ensuring that the signal will reliably arrive at its destination without the need
for specialized line conditioning.
Adaptive Protocols Another way in which errors can be prevented during data
transmission is through the use of adaptive protocols. Adaptive protocols adjust
transmission session parameters in response to varying line conditions. The MNP
Classes of networking protocols offer several examples of such adaptive protocols.
Adaptive Size Packet Assembly is an MNP Class 4 protocol that can increase or
decrease the amount of data sent in each packet according to the current condition of
the transmission circuit. A packet that includes data and overhead information is
analogous to a handwritten message (the data) plus a sealed, addressed envelope
(the overhead). This protocol tries to optimize the amount of data per packet by
building packets containing the greatest amount of data that can be transmitted reli-
ably and therefore, not require retransmission.
Optimal packet size is a moving target, changing with line conditions. Each
packet of data contains overhead. Therefore, if line conditions are good, it makes
sense to use large packets to maximize the data to overhead ration. Returning to the
previous analogy, it would be advantageous to write as long a letter as possible for
each envelope. Too little data per packet and time is wasted processing overhead
(opening envelopes in the analogy).
However, if line conditions are poor, errors will occur. Because the error detec-
tion process can only determine that an error occurred in the packet rather than the
exact bit in error, the entire packet containing the error must be retransmitted. The
larger the packets are, the more data that must be retransmitted for each error. In our
analogy this correlates to receiving, rewriting and remailing letters for insufficient
postage. Adaptive Size Packet Assembly solves the data per packet dilemma by
adapting the amount of data included in each packet according to varying line con-
ditions. When errors are detected, packet size is reduced. When no errors are
detected over a period of time, packet sizes are increased.
Dynamic Speed Shifts is an MNP Class 10 adaptive protocol that allows two
modems to change speeds up or down (faster or slower) in the midst of their data
transmission in response to varying line conditions. The adaptive nature of this pro-
tocol assures that the highest practical transmission speed will be used at all times,
dependent upon current line conditions. This adaptive protocol is especially useful
in cellular phone environments in which line quality can vary significantly over
short periods of time.
ADAPTIVE PROTOCOLS CAN MASK PROBLEMS

Dynamic Speed Shifts or any adaptive protocol can be a double-edged sword, how-
ever. In the event of degraded line quality, MNP 10 modems may automatically
downgrade their transmission speeds. Unless a personnel procedure is in place to
take note of the lower transmission speed, the problem might go undetected and
Practical Advice unreported to the carrier for an extended period of time.
and Information
Forward Error Correction Sophisticated error correction techniques exist that send
sufficient redundant data to the receiving modem to enable it to not only detect, but
also to correct data transmission errors without the need for retransmission. Forward
error correction works in a similar manner to the data correction techniques previ-
ously explained. On the transmitting side, the incoming data signal is processed and
redundant code bits are generated based on that incoming signal. These additional
redundant bits are added to the original signal and transmitted to the receiving
device. The receiving device processes the incoming data signal in the same manner
as the transmitting device and regenerates the redundant code bits. The transmitted
redundant code bits are compared with the redundant code bits regenerated by the
receiving modem. If they match, then no errors have occurred on the incoming sig-
nal. If they don’t match, the forward error correction circuitry at the receiving device
uses the transmitted redundant bits to correct the incoming data signal, thereby cor-
recting transmission errors without the need for retransmission.
However, there is a downside to forward error correction. In data communica-
tions, just as in life, you can’t get something for nothing. In order to give the receiv-
ing modem sufficient redundant data to be able to correct its own detected errors, the
overall throughput of informational data on the circuit is reduced. This process,
known as forward error correction, tries to favorably balance how much redundant
data can be sent up front, thereby avoiding retransmissions, to maximize overall
throughput on the circuit. It’s a bit of a gamble, really:
• If not enough redundant data are sent, the overall throughput is reduced due
to retransmissions.
• If too much redundant data are sent, the overall throughput is reduced
because the redundant data is taking up space and processing power that
could be occupied by “new” non-redundant data.
Trellis Coded Modulation Trellis coded modulation (TCM) is another way in

which transmission errors can be overcome without the need for retransmission.
Modems that employ TCM are able to overcome twice as much noise on a given cir-
cuit as QAM modems that don’t employ TCM. Intersymbol interference can cause a
modem to detect the wrong constellation point and subsequently interpret that con-
stellation point into an incorrect sequence of bits. Using a sophisticated technique
known as convolutional encoding, TCM adds a redundant bit that limits the possi-
ble valid constellation points for the current transmission. By limiting the number of
possible constellation points that are potentially valid for any given symbol received,
the modem is able to avoid misinterpreting symbols that would ordinarily lead to
retransmissions in order to correct the error.
For example, if we wanted to send 6 data bits per baud, we would ordinarily
require a 64-point constellation (26). However, by adding a 7th TCM code bit to each
6 data bits, a 128-point constellation would be generated (27). But remember, only 64
constellation points are required to transmit the original 6 data bits per baud, so only
64 of the 128 possible detectable constellation points are defined as valid. If one of the
invalid constellation points is detected, the TCM circuitry selects the most likely
valid constellation point and its associated pattern of bits. In this manner, TCM
reduces errors and the need for retransmission due to line impairments. However,
error detection and correction techniques must still be employed.
The sophistication of a given TCM scheme is measured by the number of poten-
tial trellis codes or states. The greater the number of TCM states, the higher the
required processing power in the modems, and the higher immunity to inter-symbol
interference due to line noise. TCM was first introduced as part of the V.32 modem
standard and is supported by V.34 and V.90 modems.
MNP4 The MNP error control standards were originally developed by Microcom, a
modem manufacturer. The MNP standards include Classes 2, 3, and 4. These error
control standards optimize the full-duplex transmission of data over dial-up lines
through Adaptive Size Packet Assembly and the elimination of redundant or over-
head information from transmissions.
V.42 Not to be confused with the CCITT V.42bis standard for data compression,
V.42 incorporates MNP Class 4 Error Control as the first of two possible error control
protocols. In addition, a second error control protocol known as Link Access Proto-
col for Modems, or LAP-M, adds Selective ARQ to the capabilities of MNP Class 4
Error Control protocol. Selective ARQ, described earlier, only requires retransmis-
sion of specific blocks received in error rather than all blocks subsequent to the block
in which the error was detected. V.42 also provides for negotiation during modem
handshaking to allow two modems to decide whether they will implement MNP 4 or
LAP-M as an error control protocol for their data transmission.
HARDWARE VS. SOFTWARE ERROR CONTROL

MNP 4 and V.42 are error control protocols implemented within modems. Using
these protocols, the modems themselves assure error-free transmission. There is no
need for additional error control protocols that can be supplied by communications
software packages running as an application program on a PC.
Practical Advice
and Information Flow Control There are two basic approaches to implementing flow control in a
modem environment: hardware and software flow control. Hardware flow control is
dependent on the use of the RS-232 serial communication standard. When using
hardware flow control the Request to Send (RTS) and Clear to Send (CTS) signals are
used to perform the essential elements of flow control. A transmitting device will
only send data into buffer memory as long as Clear to Send is “held high” (has a five
volt signal). As soon as Clear to Send is “dropped” (goes to zero volts), the transmit-
ting device immediately ceases transmitting data.
A second flow-control mechanism is to carry embed flow control flags within the
datastream itself. Special data sequences, known as flags, that mean “Stop Sending
Data” (XOFF) and “Resume Sending Data” (XON) are defined. When a device can-
not receive data any longer, it sends an XOFF signal to the sending device. When it
can resume receiving data, it sends an XON flag to the sending device.
SOFTWARE VS. HARDWARE FLOW CONTROL

When using RS-232 hardware flow control is usually more reliable and faster than
software flow control. Software flow control is nothing more than transmitted char-
acters susceptible to the same transmission problems as normal data. Occasionally,
an XOFF might be transmitted in error and might stop a data transmission session
Practical Advice for no apparent reason. Second, because XON and XOFF are sent in the datastream
and Information along with all the “normal” data and must be differentiated from that data. This
process takes time (adds latency) and is subject to error.
Application While the most ubiquitous, Internet Access Technology, dial-up

modems are the least attractive from a pure performance perspective. With best case,
real-world data speeds limited to around 45 Kbps downstream and 33 Kbps
upstream, dial-up modems are simply too slow to truly provide adequate service for
an increasingly multimedia Internet.
At this point dial-up modems have three applications in the Internet access
paradigm:
• Travel—PSTN lines are available at residences, hotels, and even modern pay
phones. When combined with a national ISP, a dial-up modem can provide
Internet access just about anywhere.
• Cost—Dial-up modems continue to be the least expensive Internet access
technology in terms of base monthly fee. However, when looking at the
cost/MB of data downloaded, dial-up trails faster technologies by a signifi-
cant margin. If a full-time connection is required, the cost of a second phone
line can quickly drive the cost of dial-up higher than faster alternatives.
• Last resort—Dial-up Internet access is available virtually everywhere, even
those places where other, faster alternatives are not yet available. In some
cases, it is simply dial-up or nothing.
Beyond the Internet access paradigm, dial-up modems have other applications,
including serving as a backup connection for other technologies such as the wide
area networking connections detailed in chapter 6. If the primary connection fails, a
dial-up connection is established to the destination to enable business to continue.
However, due to the reduced speed of a dial-up connection, care must be taken to
ensure that the limited bandwidth is used for the mission-critical applications by
preventing other, less critical applications from using the connection.
LEASED LINE BACKUP

Coping with leased-line failure requires a business decision. The technology is avail-
able to backup failed leased lines by automatically establishing connections via dial-
Managerial up lines between points on the failed circuit. However, there are at least two
Perspective incremental costs involved in this automatic backup. First, there is the cost of the
additional technology necessary to detect the leased-line failure and establish the
dial-up connection. Second, dial-up circuits must be available for the auto dial
backup unit to utilize in the event of a leased line failure. These dial-up circuits
would incur both installation and monthly charges, whether they are used or not.
In addition, if the established dial backup connection is long distance, then per-
minute usage charges for the dial-up lines could amount to a significant incremental
cost. The other side of the cost/benefit equation would require the following ques-
tion to be asked, “What is the cost of the business lost during the time that the leased
line is down?” If that lost business can be translated into lost sales dollars in excess of
the incremental cost of establishing and maintaining the dial-up backup for the failed
leased line, then the acquisition of the equipment and phone lines to enable auto-
matic backup would constitute a prudent business decision.
Digital Subscriber Line (DSL)
As the Internet has matured, so has the sophistication of the content it carries. With
Web pages growing in size due to the addition of animation and graphics, the explo-
sion of file sharing using networks such as Kazaa, and the promise of real-time voice
and video communication, dial-up modems have become too slow to meet the needs
of the Internet consumer and are being phased out in favor of “broadband” access
technologies. Broadband has many definitions, but the simplest is to consider any
access technology that provides access to the Internet at greater that 56 Kbps to be a
broadband technology.
One of the faster broadband technologies currently available is Digital Sub-
scriber Line (DSL). DSL provides an “always on” connection to the Internet over the
same copper wires that provide dial-up telephone service.
Architecture DSL uses the same physical local loop connection as the local phone
service used in POTS, as illustrated in Figure 3-15. When a local loop is used for DSL
frequency division multiplexing is used to separate the existing voice service from
the DSL service. As detailed earlier in this chapter, a POTS line uses only the bottom
4 KHz of bandwidth on the copper loop. All of the bandwidth above is available for
an alternate service, such as DSL.
As illustrated in Figure 3-21, DSL operates above the POTS frequency range. In
the figure the DSL service operates between 25 Khz and 1,100 Khz. This frequency
range is further broken into two parts for upstream and downstream data. Upstream
uses between 25 Khz and 100 Khz. Downstream data uses 110 Khz to 1100 Khz. This is
1100 kHz
Downstream
Data
110 kHz
GUARDBAND
100 kHz
Upstream
Data
25 kHz
GUARDBAND
4 kHz
0 kHz Voice
Figure 3-21 Frequency Division Multiplexing in DSL

an example of an asynchronous link. Because the amount of bandwidth given to the

downstream link is larger than that given to the upstream link, the connection sup-
ports a proportionally higher downstream data transfer speed. This type of DSL ser-
vice is commonly referred to as an Asymmetrical Digital Subscriber Line (ADSL).
Because DSL uses the same copper loop as a POTS line, it reduces the cost to the
DSL provider. If additional copper wires are needed to deliver service, the DSL
provider would have to send out a technician to install the additional wire, increas-
ing the cost of installation with a corresponding increase in DSL service cost.
In order to support DSL service, additional equipment must be installed in the cen-
tral office (CO) and the customer home (also known as the customer premises). As illus-
trated in Figure 3-22, this equipment includes a DSL modem at the customer premises
and a Digital Subscriber Line Access Multiplexer (DSLAM) at the central office.
A DSL modem is installed at the customer’s premises. This modem works like
the dial-up modems discussed earlier in the chapter in that it takes digital signals
from your computer and converts it to analog signals for transmission over the local
loop to the central office using the same digital to analog modulation technologies as
a dial-up modem. The key differences are that the DSL modem doesn’t need to “dial
up” the central office equipment every time you want to connect to the Internet: the
central office is always at the other end of the local loop connection.
When the modem is initially powered on, it communicates with the DSLAM at
the central office. The two devices then determine the optimal speed based on the
quality of the connection over the copper pair and establish a connection. This con-
nection is a point-to-point connection between the modem and the central office. The
customer is the only person using the local loop between the central office and the
customer premise. The connection remains live until the modem is powered down.
A DSL modem connects to any phone jack at the customer premise using a stan-
dard RJ-11 phone connection to receive the analog data signal from the DSLAM. The
DSL modem then un-modulates the signal and provides it to your computer (or
other network device) via a second port. The majority of the DSL modems in use
today use a standard Ethernet for this connection. The Ethernet port would be con-
nected to an Ethernet network interface card (NIC) in your PC or other network
device use as a firewall/router. Some DSL modems may use a USB connection to
PSTN
Phone
switch
Internet
Local PC DSL DSLAM Router

Local loop
Modem
Home
Central Office
Figure 3-22 DSL Architecture

connect to the computer in place of the Ethernet connection. If the modem only sup-
plies a USB connection it must be connected to a computer rather than another net-
work device.
A DSLAM is installed on the central office. This device is analogous to a modem
bank in a dial-up environment. It takes multiple incoming DSL local loops, demodu-
lates their datastreams, and multiplexes them onto a high-speed connection for
transmission to the ISP. These high-speed connections can vary from a T1 to frame
relay, to an OC-3 SONET connection. For more information about these technologies,
please refer to chapter 6. After stripping the data signal from the local loop the
DSLAM sends the underlying POTS signal to the dial-up telephone switch.
DSL modems and DSLAMs are layer two devices: they simply pass the data
between them without interpreting it. From the perspective of the user at the cus-
tomer premises the DSL modem and DSLAM are invisible.
Standards and Technology Unlike dial-up modems, there is little standardization in

the DSL world. Different vendors have developed different solutions that use differ-
ent frequencies and modulation schemes. Fortunately, the only two devices that have
to agree on the DSL technology used are the DSL modem and the DSLAM. For this
reason, most DSL service providers require customers to rent or purchase DSL
modems directly from them. This is diametrically opposite the approach taken in
cable modem environments.
There are several different DSL standards deployed. The key differences between
these technologies are the data transmission speeds they provide, the distance over
which they travel, and their reliability and resistance to noise. A matrix of commonly
available DSL services is provided in Figure 3-23.
DSL Maximum Pairs of Distance

Technology Speed copper needed Limitation Application
ADSL Up to 8 Mbps 1 Up to 18,000 ft. Consumer Internet
(Asymmetrical downstream access
DSL) 640Kbps upstream
IDSL 144 Kbps downstream 2 Up to 18,000 ft. ISDN replacement
(ISDN DSL) 144 Kbps upstream (additional and business
equipment can Internet access
extend distance)
SDSL 1.544 Mbps 1 Up to 10,000 ft. Business Internet
(Symmetrical downstream access
DSL) 1.544 Mbps upstream
HDSL 1.544 Mbps 2 12,000 to T-1 replacement
(High bit downstream 15,000 ft. and business
rate DSL) 1.544Mbps upstream Internet access
VDSL up to 34 Mbps 1 Max. of 4,000 ft. Business Internet
(Very high bit upstream and access
rate DSL) downstream,
if symmetric
Figure 3-23 DSL Comparison Matrix

Implementation and Performance A few items affect the speed of DSL services. The
first item is the copper local loop. The quality of the copper in the ground is a major
factor. If you live in an area built in the early 1950s, the copper installed in the ground
is older and might have degraded over time. If you live in a newly constructed area,
the copper could be newer and made with more modern manufacturing techniques.
The quality of the copper is something you may not be able to fix, but it is something
you need to be aware of when dealing with DSL.
Another item that affects the speed of DSL is the distance you are from the cen-
tral office. The farther you are from the central office, the longer the analog signal
needs to travel over the copper. As analog signals travel over copper over a distance,
the signal degrades in power. This signal loss affects the speed at which DSL oper-
ates. In areas where the DSL modem is close to the central office, the DSL modem
may run near its theoretical speeds. In locations where you are farther away from the
central office, the DSL modem will not operate at the theoretical speeds, but will gen-
erally operate at a speed faster than dial-up modems.
Splitter and Media Filters A splitter would be installed on the outside of the house
by the telephone company. The goal is to separate the data service from the voice ser-
vice. This was needed due to the nature of how phone service works. When a phone
goes off hook, a power spike is caused by the contacts opening up on the phone
switch. The power spike sends a frequency blip over the entire frequency spectrum
allowed to pass over the copper pair. The frequency blip, if not stopped, would cause
the DSL modem to “momentarily” drop connection with the DSLAM in the central
office. This would cause data to be lost and the modem would have to reestablish
their connection.
Early DSL modems required these splitters to be installed to eliminate the possi-
bility of phone calls affecting the data service. The splitter separated the wiring
inside the house to a phone set of connections and a single data connection. The
phone connection also had a device installed to eliminate the power spike. The
installation of the splitter increased the cost of the service due to the labor of the per-
son from the telephone company being required to visit each residence to install the
splitter. A consumer installation method was required that would eliminate the truck
roll, thus lowering the cost and decreasing the time it took to turn up each service.
The solution was found in the use of media filter.
Media filters are sent with the DSL modem and installed by the consumer.
Media filters are wire-wrapped magnets that are installed on each phone in the
house that is connected to the phone line the DSL is on. The goal of the filter is to
suppress the high-voltage spike that occurs when a traditional phone is taken “off
hook.” Media filters absorb the frequency blip and allow the DSL modem to work
without interruption.
Consumer Class vs. Business Class DSL As discussed above, DSL can be used for both
business and consumer access to the Internet. As you would expect, business-class
DSL cost more than consumer-class service. Consumer-class DSL usually uses DHCP
for IP addressing and usually allocates 1 IP address per DSL modem. The bandwidth
allocation is generally more in the downstream direction and the upstream direction.
This matches the wants of most consumers: they need to get more information from
the Internet than they send.
Business-class DSL generally allocates a static IP address to the business.
Additional IP addresses can be purchased, for a cost. Businesses generally want to
run a Web site from their location, so the data requirements are different than the
“standard” consumer. Bandwidth allocation is generally more symmetrically with

speed moving toward T1 speeds (1.544Mbps upstream and downstream). The
business might also want the ISP to supply e-mail services for all their employees.
This will also increase the cost.
If a consumer needs some of the features that a business would need, the
providers are more than happy to sell them a business class DSL connection. The
only catch is that they will pay the same cost as the business would. Some providers
do offer a more cost-effective consumer solution that deletes some of the other key
features that a business user would need.
Future of DSL The future of DSL is focused on providing new services over the DSL
modem. DSL providers would like to provide audio and video-on-demand services
over the DSL modem. Both of these services will require the need for quality of ser-
vice (QoS) to allow them to work effectively. QoS will allow the user to make voice
calls over the Internet, stream video to PC, and download files at the same time with-
out the voice or video quality suffering from the file download.
Cable Modems
A second provider of high bandwidth connectivity to customer premises is the tele-

vision cable company. In fact, the cable provider’s infrastructure offers a significantly
higher bandwidth to the consumer than the local loop provided by the telephone
company due to the coaxial cable media used for cable television transmission.
Architecture Cable systems were originally designed to provide one-way communi-

cation between the programming source, known as the head end, to the consumer’s
television set. Analog television programming was broadcast across the network to
any and all television receivers connected to the system via coaxial cable. Splitters
and amplifiers were placed in the system where needed to allow the system to grow
to meet the needs of the community served. This type of cable television network
looks like a tree: the head end is the root and the branches continually are split off
and fed into neighborhoods.
Modern cable networks have evolved to support two-way communication in
addition to one-way, broadcast communication with increased bandwidth available.
The increased bandwidth can be used by the cable company to offer a larger number
of television channels in either analog or digital form, offer on-demand television
where a customer can request movies to be sent directly to their home, and offer
interactive Internet access.
To offer bi-directional Internet access, frequency division multiplexing is used to
reallocate the bandwidth for one television channel (in the 50 to 750 MHz range) for
downstream traffic and another channel (in the 5 to 42 MHz band) is reallocated to carry
upstream signals. A single downstream 6-MHz television channel can support up to 27
Mbps of downstream data throughput using 64 QAM (quadrature amplitude modula-
tion) transmission technology. Speeds can be boosted to 36 Mbps using 256 QAM.
Upstream channels may deliver 500 Kbps to 10 Mbps from homes using 16QAM or
QPSK (quadrature phase shift key) modulation techniques, depending on the amount
of spectrum (number of channels) allocated for service and the relative noise on the line.
Once the bandwidth to be used for Internet access has been allocated, equip-
ment must be installed at the head end and at the customer premises. Cable modem
service providers use a Cable Modem Termination System (CMTS) to add data
transmission capabilities to the existing CTV infrastructure. The CMTS is capable of
providing throughput to multiple homes at rates of up to 50 Mbps. However,
throughput is usually less than 27 Mbps due to the way data are transmitted to
homes from the cable modem service provider.
As can be seen in Figure 3-24 the CTMS is capable of serving multiple distribu-
tion hubs, which, in turn, serve multiple fiber nodes. Each fiber node serves a neigh-
borhood, or physical grouping of homes. Most groupings range from 500 to 2,000
homes on a modern, HFC network. This architecture typically results in speeds to the
home of between 500 Kbps and 1.5 Mbps, depending on network congestion.
Between the fiber nodes and the homes in the neighborhood, the upstream and
downstream bandwidth is shared by the active data subscribers connected to a given
cable network segment. A device known as a diplexer is used to both separate and
combine the upstream and downstream signals and is located at the fiber node.
Once the coaxial cable reaches the subscriber’s home, a frequency splitter is
used to separate the TV signals from the data signals. Assuming the home is pre-
wired for CTV, the existing wiring is used for the CTV signals. The “new” wire is
connected to a cable modem, which, in turn, connects to the subscriber’s PC via 10
Mbps Ethernet or, optional in newer cable modems, a USB connection. Figure 3-25
illustrates this configuration.
The function of a cable modem is essentially the reverse of the CMTS: to modu-
late and demodulate upstream and downstream signals, respectively. As mentioned,
the CMTS uses the same techniques on the upstream and downstream signals at the
head end. Downstream data is modulated at the headend using the 64-QAM modu-
lation technique. The data occupies a 6-MHz channel somewhere in the 50 to 750
MHz range. The signal is demodulated at the cable modem side.
Upstream signals are handled differently. Like their downstream counterparts,
they are modulated at the cable modem side and transmitted back to the CMTS.
Because multiple cable modems must share the upstream bandwidth to the CMTS, a
distribution
Internet fiber/copper
hub
node
622 Mbps distribution

(OC - 12) hub fiber/copper
node
IP over SONET,
ATM, or WDM
CMTS
Cable
HeadEnd
distribution fiber/copper
hub node
Figure 3-24 Cable Data Network Architecture

Coax
VCR TV
Splitter
HOME
Coax
Coax
TV
Splitter
Coax Cable Coax Cable Modem 10BaseT or Computer

USB
Figure 3-25 Typical Cable Modem Installation
multiple access multiplexing scheme must be used. Time Division Multiple Access
(TDMA) and Frequency Division Multiple Access (FDMA) are commonly used to
allow multiple cable modems to communicate to the CMTS concurrently. Before a
cable modem can transmit to the CMTS, it must request and receive a bandwidth
allocation. This is usually handled as part of the initial power-on handshaking
between the cable modem and the CMTS.
Standards and Technology Initially, there were no standards that defined the modula-
tion schemes used between cable modems and the CMTS. As a result, cable modem
customers were locked into the cable modems offered by their cable company. In
1996, several cable operators commissioned the development of a standard, which
would control compatibility of cable modem equipment, both on the user side, the
cable company side, and the wiring in between. The resulting Data Over Cable Ser-
vice Interface Specification (DOCSIS) standard defines the protocol stack used in
transmissions, modulation techniques for the CMTS, and cable modem and line
speeds. Both cable modems and CMTS systems can be certified as DOCSIS-compli-
ant, which indicates all aspects of the specification are adhered to and ensures inter-
operability between vendors. As long as the CMTS is DOCSIS compliant, the
consumer can purchase any DOCSIS-compliant cable modem with certainty that it
will work with their cable system.
The DOCSIS standard also includes some security enhancements for the privacy
of data. This basic privacy specification is called the BPI, or Baseline Privacy Inter-
face. BPI allows for the use of encryption algorithms to assure the privacy of data
when traveling between the user’s cable modem and the cable company. It should be
noted, however, that the BPI was developed to benefit the provider, not the user. The
implementation of the encryption is done at the cable company, meaning subscribers
have no control over whether encryption is activated or not for their shared “ring.”
When activated, encryption provides several benefits, including authentication of

users’ modems (preventing theft of service), and virtually eliminates the sniffing
threat between the cable modem and the CMTS at the cable company.
An extension to the BPI, BPI+, extends the capabilities of BPI to eliminate cloning
of cable modems. Prior to the implementation of BPI+ it was possible for software or
hardware to clone a cable modem and have it communicate successfully to a CMTS.
BPI+ allows increased authentication by incorporating the cable modem’s physical
address into the encryption scheme.
Implementation and Performance In many ways a cable modem system has more in
common with an Ethernet local area network than it does with a dial-up modem or
DSL system. Unlike dial-up or DSL, where the bandwidth between the customer
premises and the central office is dedicated to the customer, cable modem systems
require that everyone on the cable segment share a fixed amount of bandwidth.
Because of this, users might find that the service that was blazingly fast when only a
few people shared the segment has grown increasingly slower as more of their
neighbors connect. This shared bandwidth scenario also has security ramifications:
the downstream traffic destined for one cable modem is received by all cable
modems on the segment. Although each cable modem normally ignores traffic not
destined for it, if placed in promiscuous mode the cable modem will pass all incom-
ing traffic to the attached computer. For more information about broadcast shared
media networks, please refer to chapter 4.
Cable modems themselves come is three forms: external, internal, and set-top
box. External cable modems, by far the most popular, are physically distinct. Like all
cable modems, they translate the data signal between the coaxial cable that the cable
company supplies and the 10Base-T Ethernet signal the user’s PC expects. This fla-
vor of modem allows for multiple PCs to share the connection but requires a separate
Ethernet network card to be installed in the subscriber’s PC.
Internal cable modems are packaged as a PCI card installed in the subscriber’s
PC. Internal cable modems are less expensive than their external counterpart, but
they are not “plug” compatible with laptop computers and Macintosh machines.
The final cable modem form factor is the set-top box, which provides network
access on the subscriber’s television. Set-top implementations require a separate
upstream method, usually a standard dial-up modem for return signals. Typically,
this type of implementation was only used for older, broadcast-only cable systems
rather than modern bi-directional systems.
Future of Cable Modems Cable modems offer the highest Internet access connection
speeds currently available. Although the shared media approach used in cable
modems requires the bandwidth on an individual cable segment to be shared among
the users on the segment, most cable modem customers receive excellent data
speeds. As cable companies gain experience in the data communications market-
place, they are in an excellent position to leverage their existing cable plans to pro-
vide additional interactive services to consumers.
Other Internet Access Technologies

Several other Internet access technologies are under development. One of the most
interesting is fixed-point wireless. In a fixed-point wireless environment, the service
provider places a series of transmitters on a tower. Each transmitter is directional in

nature and covers a portion of the 360 degrees around the tower. Consumers place a
receiver unit in their home that connects to the tower via radio waves. Depending on
their distance from the tower, consumers may need a directional external antenna to
establish a radio link. Once the link is established, the consumer can access the Internet.
There are several different wireless access schemes under development, ranging
from variants of IEEE 802.11 wireless LANs to licensed spectrum solutions. For more
information on IEEE 802.11 networks, please refer to chapter 4.
Originally thought to best fit into urban/suburban settings, wireless Internet
access is instead making its initial market penetration in rural areas. Upon closer
inspection, rural areas are perfect for wireless access: there are few tall buildings to
create multi-path interference in the radio signal and there are typically few compet-
ing broadband access technologies because cable and DSL providers have been focus-
ing on more densely populated areas. Some smaller rural towns have begun
proactively seeking wireless access providers to remain competitive at attracting new
residents to their tax base. It is becoming increasingly common to see an antenna
array installed on the top of grain elevators in small towns across the country.
Cellular telephone companies are working to develop faster Internet access tech-
nologies to serve their mobile customers. Although the cost of these services is cur-
rently too high to make them a viable choice for non-mobile users, it is too soon to
rule them out as potential players in the market. These technologies are discussed in
detail in chapter 5.
There are many different wide area network services available that can be used
to provide Internet access. While these solutions are normally considerably more
expensive than the previously discussed solutions, they offer the advantage of being
dedicated to your use. A complete discussion of wide area network technologies is
presented in chapter 6.
INTERNET ACCESS TECHNOLOGY SELECTION FRAMEWORK

There is no one “best” Internet access technology solution for everyone. To help
decide the best solution for a given situation the following series of questions should
be answered:
Applied Problem
Solving • What types of service are available at the location? There may be only one
choice available.
• How much bandwidth is required/desired?
• Which of the available services can meet the bandwidth requirements?
• Will Internet services such as the Web and e-mail be hosted on premises?
❍ If so, an “always on” broadband service such as DSL, cable, or a wide area
network solution is required.
❍ Be sure to check the usage rules for any service considered. Some cable
companies, for instance, have chosen to block incoming Web traffic to pre-
vent users from running their own Web servers.
After analyzing your needs and the available solutions, it is not uncommon to
find a clear winner. For those people fortunate enough to have multiple viable solu-
tions, monthly cost and data speeds should be the deciding factors.
Key Terms 111
SUMMARY
By focusing on how data communications Current modulation standards such as V.34

devices of any type deliver on pre-determined and V.90 can deliver even more throughput over
business requirements and objectives, one can dial-up lines when compression standards such as
avoid purchasing technology that may be MNP5 or V.42bis are applied. As this increasing
appealing or cleverly marketed, but which lacks sophistication in dial-up modems has yielded
the ability to deliver a positive impact on busi- ever faster transmission speeds over dial-up lines,
ness objectives. the types of network services offered by carriers
LAN media can differ significantly in cost, have also evolved.
supported speeds, ease of use, and network archi- Regardless of the modem technology used
tectures supported. Although fiber optic cable dial-up data calls have limited data carrying
was at one time considered to be the only media capacity. To access the power of the Internet a
suitable for speeds of 100Mbps and greater, Cate- faster means of connecting is required. Often
gory 5 Unshielded Twisted Pair seems to be a referred to as broadband, these faster connections
common media option for high speed network currently use two basic technologies: DSL and
standards. New advanced testing techniques have Cable modems.
been developed to assure that UTP will be capa- Digital subscriber line (DSL) technologies use
ble of transmitting high speed network architec- frequency division multiplexing to add a high
tures such as Gigabit Ethernet. speed data signal to the existing telephone local
Point-to-point data communication occurs in loop. The data is modulated and multiplexed in a
either a serial or parallel manner. Modern connec- DSL modem that is connected directly to the tele-
tions are almost exclusively serial in nature. Com- phone wiring in the home. At the central office
mon standards include RS-232, universal serial the signals are de-multiplexed by a DSL access
bus (USB) and IEEE-1394 (Firewire). multiplexer (DSLAM). DSL offers data speed in
Using the transmission of data between a excess of 1 Mbps over an existing direct connec-
computer and the Internet as an example of a sim- tion to the home.
ple data communications system, many key con- Cable modem technologies use frequency divi-
cepts from last chapter were show in context in sion multiplexing to add high a high speed data
this chapter. These technologies include dial-up signal to the existing cable television cable coming
modems, digital subscriber line modems, and into the home. Similar to DSL, a cable modem han-
cable modems. dles the modulation/demodulation at the home
Current modem standards include V.34 and and a cable modem termination system (CMTS)
V.90. V.34 is the newest analog modem standard handles it at the cable head end. However, unlike
offering transmission speeds up to 33.6 Kbps in DSL the cable modem communication channel is
both directions. V.90 is a hybrid analog/digital shared among all of the users on the cable node.
standard that offers an asymmetrical transmission New Internet access technologies such as
rate of up to 56 Kbps from a remote server to the fixed point wireless and 3G digital cellular sys-
modem and up to 33.6 Kbps from the modem to tems promise to further improve the availability
the server. of high speed data connectivity to the home.
KEY TERMS
physical layer UTP Multimode

four conductor station wire EIA/TIA 568 Multimode Step Index Fiber
RYGB Attenuation Multimode Graded Index Fiber
flat gray modular Near-End Crosstalk (NExT) single mode
unshielded twisted pair Powersum Crosstalk DTE (data terminal equipment)
DCE (data communications Bandwidth Forward error correction

equipment) Handshaking Trellis coded modulation (TCM)
data circuit terminating equipment Turnaround time Convolutional encoding
UART (Universal Asynchronous Echo cancellation Link access protocol for modems
Receiver Transmitter) Digital signal processors (LAP-M)
universal serial bus (USB) DSP Broadband
IEEE-1394 V.90 Digital subscriber line (DSL)
Firewire V.34 Asymmetrical digital subscriber
i-Link V.42 line (ADSL)
IrDA Quantization noise Digital subscriber line access
Bluetooth Asymmetrical transmission multiplexer (DSLAM)
ISP V.42 bis Splitter
Internet Service Provider MNP Class 5 Media filter
POTS Repeater Cable modem termination system
Plain Old Telephone Systems Amplifier (CMTS)
Public Switched Telephone Adaptive protocols Diplexer
Network Adaptice size packet assembly DOCSIS
PSTN MNP Class 4 Data over cable service interface
Dial-up line Dynamic speed shifts specification
Circuit-switched connection MNP Class 10 Fixed point wireless
REVIEW QUESTIONS
1. How can the top down model remain useful given 17. Differentiate between IEEE-1394, Firewire, and
the rate of rapidly changing technology? i.Link.
2. Why is twisted pair twisted? 18. At what speeds does IEEE-1394 operate?
3. What is the importance of EIA/TIA 568? 19. What is currently the primary application for
4. What is the most common type category of UTP IEEE-1394?
installed today and why? 20. List the currently available versions of IEEE-1394
5. Why is UTP category 5 favored over shielded and the maximum data speeds offered by each.
twisted pair, coax, and fiber optic cable for many 21. Which point-to-point serial transmission stan-
high speed network architectures? dards offer power to attached devices?
6. Why is shielded twisted pair considered trickier to 22. What is modulated to provide wireless connectiv-
install than UTP? ity via IrDA?
7. What is the difference between powersum 23. What are some of the significant limitations associ-
crosstalk and pair-to-pair crosstalk? ated with IrDA?
8. How are Levels 5, 6, and 7 different from Cate- 24. Which IrDA standard offers multi-point communi-
gory 5? cation?
9. Why are testing and certification specifications 25. What is the size and data transmission speed
required beyond CAT 5? offered by Bluetooth?
10. What are the key advantages of fiber optic media 26. What type of radio frequency communication
over copper media? does Bluetooth use?
11. List the two main types of fiber optic media and 27. What is a piconet?
the advantages of each. 28. What types of devices are likely to include Blue-
12. List the 10 most commonly used RS-232 pins tooth capability?
including name, abbreviation, and DCE/DTE ori- 29. What is the purpose of the I-P-O model and how
entation. can it be used to model both ends of given circuit?
13. What is the name of the device employed to moni- 30. What is an Internet service provider (ISP)? What
tor and manipulate RS-232 signals? types of services does an ISP provide?
14. At what speeds does USB operate? 31. What is the potential throughput of a V.34 modem
15. List three advantages of USB over RS-232. with V.42bis data compression?
16. List the currently available versions of USB and 32. Why are standards important when it comes to
the maximum data speeds offered by each. data compression?
Review Questions 113
33. What is meant by an adaptive protocol? Give at 51. What are the maximum speeds and distance limita-
least two examples. tions of the most commonly available DSL services?
34. What are the key differences between V.32 and 52. Why are splitters or media filters required in DSL
V.32bis? implementations?
35. What does “bis” stand for? 53. What is the difference between consumer and
36. What is the difference between V.42 and V.42bis? business class DSL offerings?
37. How does a V.90 modem system overcome the 54. Is DSL a shared or dedicated connection into the
33.6 Kbps limit as dictated by Shannon’s Law? home?
38. What is quantization error? 55. What type of multiplexing is used by a cable
39. Explain why V.90 is asymmetrical in nature? modem?
40. How fast can a V.90, 56 Kbps modem actually 56. In a cable modem deployment how many televi-
transfer data? sion channels are eliminated to provide bi-direc-
41. Will a V.90 modem provide faster speeds than a tional data services?
V.34 modem in all cases? Why or why not? 57. What piece of equipment is required at the cable
42. What are the major differences between repeaters system head end to enable cable modem usage?
and amplifiers? 58. Is cable modem access a shared or dedicated con-
43. Explain the differences between hardware and nection to the home?
software flow control. 59. What is a cable node and what is its relevance to
44. What is forward error correction and what is the cable modem services?
trade-off involved in such a protocol? 60. What is a diplexer and what is its purpose in a
45. Differentiate between MNP5 and V.42bis. cable modem installation?
46. Differentiate between MNP4 and V.42 61. What is DOCSIS and why is it important?
47. What are the main applications of dial-up Internet 62. What are the technologies developed to improve
access today? security in a cable modem deployment?
48. What does the term broadband mean in the con- 63. What are the three most common form factors for
text of Internet access? a cable modem? List the advantages of each.
49. What existing communication link into the home 64. Which offers higher data rates: DSL or cable
does DSL use? modems?
50. What piece of equipment is required at the central 65. What is fixed-point wireless and where is it ini-
office to enable DSL use? tially being deployed

Basic Communication Technology

Uploaded by

Copyright:

Available Formats

Basic Communication Technology

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Basic Communication Technology

Uploaded by

Copyright:

Available Formats

CHAPTER 3

BASIC DATA COMMUNICATION

1. Understand the types of communication media available and the limitations

Not Twisted Pair

Unshielded Twisted Pair

Although category 1 UTP, otherwise known as voice-grade, need only carry

UTP Maximum Data Attenuation

Figure 3-1 Unshielded Twisted Pair Specifications

• Attenuation is the decrease in the power of signal over a distance in a partic-

Highest Test Required Frequency in MHz for Attenuation to Crosstalk Ratio

Figure 3-2 UTP Level 5, 6, and 7 Performance Specifications

COMMON UTP INSTALLATION MISTAKES

STP-Shielded Twisted Pair

Solid metal Plastic insulator Foil shield Braided Plastic or vinyl

Figure 3-3 Coax Cable: Cross-Section

fiber of glass, it is immune to electro-magnetic interference, contributing to its high

Glass core Glass cladding Plastic or vinyl jacket

50 microns 125 microns

62 microns Multimode 125 microns

Note: A micron is a millionth of a meter

Figure 3-4 Fiber-Optic Cable: Cross-Section

Unshielded UTP 100 Mbps 100 feet RJ-45 5 Designated categories.

Shielded STP 16 Mbps 100 feet RJ-45 or IBM Shielding reduces

Fiber-optic Fiber several several SI or SMA 905 Difficult to install but

Figure 3-5 LAN Media Technology Analysis

■ POINT-TO-POINT DATA TRANSMISSION TECHNOLOGIES

Serial Transmission Standards

Serial transmission is the basis of most data communication between computers.

Pin Pin Pin

1 Protective ground 14 Secondary transmit data 1 Carrier detect

2 Transmit data 15 Transmit clock (DCE) 2 Receive data

3 Receive data 16 Secondary receive data 3 Transmit data

4 Request to send 17 Receiver clock 4 Data terminal ready

5 Clear to send 18 Receiver dibit clock 5 Protective ground

6 Data set ready 19 Secondary request to send 6 Data set ready

7 Signal ground 20 Data terminal ready 7 Request to send

8 Carrier detect 21 Signal quality detector 8 Clear to send

9 Positive DC test voltage 22 Ring indicator 9 Ring indicator

10 Negative DC test voltage 23 Data signal rate selector

11 unassigned 24 Transmit clock (DTE)

12 Secondary carrier detect 25 Busy

13 Secondary clear to send

PHYSICAL INTERFACES VS. TRANSMISSION PROTOCOLS

DB-25 connector (female) DB-9 connector (female)

Figure 3-7 Common RS-232 Connectors

RS-232 DB- DB-

Figure 3-8 Most Commonly Used RS-232 Signals

Serial/Parallel Conversion: UARTs Data travel via parallel transmission within a PC

Figure 3-9 Other Traditional Serial Transmission Standards

OTHER TRADITIONAL SERIAL TRANSMISSION STANDARDS

Figure 3-10 USB Hub Implementations

To reduce the configuration hassles traditionally associated with serial connectivity

IEEE-1394/Firewire Another high-speed serial standard is IEEE 1394. Originally

USB-A (on left) and USB-B (on right) connectors

Figure 3-11 USB Connectors

Parallel transmission is primarily limited to transmission of data within a computer,

Figure 3-12 IEEE-1394 Connectors

DB-25 (male) parallel interface Centronics parallel interface

Figure 3-13 Parallel Transmission Physical Interfaces

Infrared (IrDA) The Infrared Data Association (IrDA) is an international standards