Professional Documents
Culture Documents
McGraw-Hill
New York Chicago San Francisco Lisbon London Madrid
Mexico City Milan New Delhi San Juan Seoul
Singapore Sydney Toronto
McGraw-Hill
abc
Copyright © 2002 by The McGraw-Hill Companies, Inc. All rights reserved. Manufactured
in the United States of America. Except as permitted under the United States Copyright
Act of 1976, no part of this publication may be reproduced or distributed in any form or by
any means, or stored in a database or retrieval system, without the prior written permission
of the publisher.
ISBN 0-07-138247-X
All trademarks are trademarks of their respective owners. Rather than put a trademark symbol
after every occurrence of a trademarked name, we use names in an editorial fashion only,
and to the benefit of the trademark owner, with no intention of infringement of the trademark.
Where such designations appear in this book, they have been printed with initial caps.
McGraw-Hill eBooks are available at special quantity discounts to use as premiums and sales
promotions, or for use in corporate training programs. For more information, please contact
George Hoare, Special Sales, at george_hoare@mcgraw-hill.com or (212) 904-4069.
TERMS OF USE
This is a copyrighted work and The McGraw-Hill Companies, Inc. (“McGraw-Hill”) and its
licensors reserve all rights in and to the work. Use of this work is subject to these terms.
Except as permitted under the Copyright Act of 1976 and the right to store and retrieve one
copy of the work, you may not decompile, disassemble, reverse engineer, reproduce, modify,
create derivative works based upon, transmit, distribute, disseminate, sell, publish or subli-
cense the work or any part of it without McGraw-Hill’s prior consent. You may use the work
for your own noncommercial and personal use; any other use of the work is strictly prohibited.
Your right to use the work may be terminated if you fail to comply with these terms.
THE WORK IS PROVIDED “AS IS.” McGRAW-HILL AND ITS LICENSORS MAKE
NO GUARANTEES OR WARRANTIES AS TO THE ACCURACY, ADEQUACY OR
COMPLETENESS OF OR RESULTS TO BE OBTAINED FROM USING THE WORK,
INCLUDING ANY INFORMATION THAT CAN BE ACCESSED THROUGH THE
WORK VIA HYPERLINK OR OTHERWISE, AND EXPRESSLY DISCLAIM ANY
WARRANTY, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO
IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR
PURPOSE. McGraw-Hill and its licensors do not warrant or guarantee that the functions con-
tained in the work will meet your requirements or that its operation will be uninterrupted or
error free. Neither McGraw-Hill nor its licensors shall be liable to you or anyone else for any
inaccuracy, error or omission, regardless of cause, in the work or for any damages resulting
therefrom. McGraw-Hill has no responsibility for the content of any information accessed
through the work. Under no circumstances shall McGraw-Hill and/or its licensors be liable for
any indirect, incidental, special, punitive, consequential or similar damages that result from
the use of or inability to use the work, even if any of them has been advised of the possibility
of such damages. This limitation of liability shall apply to any claim or cause whatsoever
whether such claim or cause arises in contract, tort or otherwise.
DOI 10.1036/007138247X
Contents
Copyright © 2002 by The McGraw-Hill Companies, Inc. Click here for Terms of Use.
2 Access Technologies: DSL and Cable
Local trunks
Y
Access trunks
Telephone
FL
SONET/SDH Backbone
Local Switch Toll Switch
Telephone Local trunks
6
AM
TE Local Switch
PBX
Local
Customer Premise Local Local Toll Transmission
Access
Equipment Loop Switching Switch Equipment
Trunks
Splitter
Splitter
Feeder
11
Cables
Subscriber
Drops Taps
Wireless Access
Wireless has long played a vital role in telecommunications,
primarily in the long-haul transmission network and in
mobile applications. Satellite services once were—and for
some countries still are—a way of bringing telecommunica-
tions services to locales that lack heavy enough demand to
justify undersea cable. Point-to-point terrestrial microwave
was once the primary means of intercontinental telecommu-
nications service, and still is a convenient way of bypassing
obstructions that are expensive to cross with cable. In the
access market, however, wireless has been impaired by sev-
eral factors. First is the fact that broadband demand didn’t
really develop in the local network until fiber optics was
available and the quality of fiber is so much higher that is
always preferred if it is economically feasible.
The most limiting aspect of wireless has been the lack of
available frequencies. Microwave got its start from radar
technologies that were first employed in World War II and
then converted to commercial telecommunications service
Introduction to Access Technologies 13
Copyright © 2002 by The McGraw-Hill Companies, Inc. Click here for Terms of Use.
16 Access Technologies: DSL and Cable
Modulation Methods
Data is passed across a network by having its signal encoded
on the transmission medium through a process known as mod-
ulation. Baseband modulation applies the signal directly to the
physical facility. A telephone system, for example, applies the
Data Transmission Fundamentals 17
T1/E1 Carrier
The basic digital facility over which virtually all voice and
data signals pass is known as T1/E1 multiplex or carrier.
Data signals can be connected directly to a digital circuit
through a device known as a channel service unit (CSU).
Analog signals are digitized through a process known as
pulse code modulation (PCM). Nyquist’s law states that if the
amplitude of an analog signal is sampled at twice the highest
frequency it contains, the samples can be used to reconstruct
the original signal with a reasonable degree of fidelity. Since
the highest frequency in the audio pass band is 4 kHz, this
means that a telephone signal is sampled 8000 times per sec-
ond. Each sample is scaled into an 8-bit word through a
process called quantizing. This signal, which is known as a
DS-0 in North America, comprises a 64-kbps signal, which
is derived from the product of 8000 samples per second × 8
bits per sample. In North America, 24 DS-0s are time-divi-
sion multiplexed into a T1 frame. The E1 frame in Europe
consists of 32 circuits, of which two are used for signaling
and 30 for information. Although the T1/E1 signal is chan-
nelized into 24 or 30 channels for voice, it can be obtained
unchannelized for wideband data. In ISDN terms, a
T1/E1signal is known as a PRI.
PCM is not a new process. It was developed in England in
1938, but it wasn’t practical with vacuum tube technology
because of the problems of power drain and need for floor
space. A decade later, the invention of the transistor brought
PCM to the realm of practicality, but not until the 1960s did
20 Access Technologies: DSL and Cable
Superimposed
Sine Wave
0
0
10
1
00
0
01
1
11
1
Level 2 +
Level 1 +
Level 1 -
Level 2 -
+
-
Unipolar
Bipolar
2B1Q
Data Protocols
Data devices communicate with one another by using pro-
tocols, which are sets of programmed instructions that a
processor can execute. Protocols are used to set up sessions
between devices, determining such factors as which end con-
trols the session, what participants are authorized, what
transmission speed and code set will be used, and myriad
other functions. Protocols handle other such functions as
addressing, error detection and correction, and recovery from
network failures. Before devices can communicate, they
24 Access Technologies: DSL and Cable
must confirm that they are prepared to use the same protocols
and resolve how they will apply any optional features. Data
protocols are classified as connection-oriented or connection-
less. In a connection-oriented protocol the path is established
at session setup and remains for the duration of the session.
The telephone network is a prime example. A connectionless
protocol launches data packets into the network, where they
are routed to the destination as an independent unit. Internet
Protocol (IP) is a connectionless protocol.
As with most computer programs, protocols are built in
modules or layers. Each module has a specific function and
has clearly identified application programming interfaces
(APIs) at the boundary so developers can write to and use the
functions of the protocol. Table 2.2 shows the layers and their
functions in the International Standards Organization (ISO)
Open Systems Interconnect (OSI) model. OSI is not intended
to be a complete protocol. Instead, it is a conceptual model
with descriptions of the functions that fit within each layer so
developers can build on its structure.
Table 2.2 shows a brief summary of the functions of each
layer in OSI. Each end of the session uses the same layer
definitions. The sending end of the session passes information
to the receiving end by appending records to the information
block. The receiving end receives the instructions in the block,
executes them, and strips the extra records until only the infor-
LAN Protocols
The operation of local area network (LAN) protocols illus-
trates in practical terms how the OSI model functions. This
will also aid in understanding the operation of 802.11b wire-
less, discussed in Chap. 7. The IEEE 802 Committee, which
developed the LAN protocols, designed the protocols to
operate within the first two layers of the OSI model. The
most popular LAN protocol, Ethernet, is used almost univer-
sally as a LAN protocol to connect to the Internet. Figure 2.3
shows the layered structure of the Ethernet protocol stack.
Ethernet drivers are available to connect to any of the popu-
lar transmission media. Early implementations used RG-8
and RG-58 coaxial cable, but unshielded twisted-pair wire
(UTP) is by far the most common medium. Drivers are also
available for fiber optics and wireless systems.
Ethernet’s link layer is divided into two portions; media
access control (MAC) and logical link control (LLC). The
MAC layer is responsible for controlling access to the medi-
um—which, in the case of Ethernet, is based on contention.
A station with information to transmit listens to the network
to see if it is idle. When it determines that no other station is
sending, it launches an Ethernet frame, which is shown in
26 Access Technologies: DSL and Cable
Higher Layer
Protocols
Logical
Link
Control
LLC
Data Link Layer Media
Access
Control
MAC
Physical
Link
Signaling
PLS
Physical Layer
Attachment Unit Interface
Destination Source
Preamble Length Data Pad CRC
(1) Address Address
(7) (2) (0-1500) (0-46) (4)
(2 or 6) (2 or 6)
Routing
Since a bridge is a two-port device, it lacks the ability to
make routing decisions. Some bridges may deliberately or
inadvertently develop more than one path between segments.
This condition is prohibited by the Ethernet protocol, and is
prevented by using a protocol known as spanning tree. More
frequently, a routing protocol such as IP is used and the net-
work is connected through routers. The TCP/IP protocol falls
under the blanket of packet switching. The U.S. Department
of Defense initially conceived packet switching as a security
measure. If a message is broken into small packets and sent
through the core network over random patterns, it is difficult
to reassemble the message except at the access point. If the
access circuit is kept short, it is easy to secure. Furthermore,
if the core network has plenty of bandwidth, packets from
other sessions can be interleaved to use the excess capacity.
The problem with a connectionless network is that packets
can arrive with errors or out of sequence, and some method
must be used to preserve the integrity. Packet reassembly is
one function of data protocols such as TCP.
IP operates on top of the Ethernet LLC at layer 3 in the
OSI model, and introduces a second layer of addressing. The
MAC address is always the means by which a station can be
identified. It is permanently burned into the network inter-
Data Transmission Fundamentals 29
Application Services
TELNET
Transmission User
Control Defined
Protocol Protocol
(TCP) (UDP)
Internet Protocol
(IP)
Address Reverse
Resolution Address
ARP RARP
Protocol Resolution
Protocol
Datalink
Physical
Frame Relay
Toward the end of the 1980s, a new protocol known as frame
relay began getting attention. Unlike fixed circuits, frame relay
34 Access Technologies: DSL and Cable
Copyright © 2002 by The McGraw-Hill Companies, Inc. Click here for Terms of Use.
38 Access Technologies: DSL and Cable
Serving
Area Subscriber
Interface Drops
Distribution
Cable
Feeder Cables
Central Office
Serving
Area
Interface
Distribution
Cable
Subscriber
Drops
C
nc
ap
ta
uc
ac
d
ita
In
nc
e
Frequency
Reactance
Figure 3-2 Reactive Effects of Capacitance and Inductance
Crosstalk Considerations
The local loop was designed and constructed to support
voice-frequency signals. The use of one pair to serve one
subscriber does not begin to use the bandwidth capacity of a
cable pair. For short distances, plain copper wire is cost-
effective, but for longer subscriber loops or for large concen-
trations of loops into a single subscriber, multiplexing can
increase the capacity. Therefore, the ILECs sometimes use
multiplex equipment in the local loop. Several factors insti-
gated this. One was the development of digital loop carrier
(DLC), which allows the ILECs to extend subscriber lines to
the outer reaches of the wire center using a system similar
to T1/E1. The DLC is housed in an environmentally con-
trolled enclosure and fed from the CO with fiber optics. This
is a cost-effective alternative to using coarser cable gauges
44 Access Technologies: DSL and Cable
Near End
Transmitter
Cable Pairs
Far End
Receiver
45
Far End
Transmitter
Echo Considerations
Another impairment that must be considered is echo. Echo
occurs whenever a four-wire circuit is converted to a two-wire
circuit. Trunks are almost always four-wire because they are
provisioned on multiplex equipment that inherently has sepa-
rate transmit and receive paths. Local loops were designed
as two-wire to conserve cable. Many older switches were also
two-wire, but all digital switches are four-wire, meaning the
Y
transmit and receive directions of a session are separated. A
FL
device known as a hybrid does the four-wire-to-two-wire con-
version. A hybrid works on the principle of balancing the
impedances of the two-wire and four-wire loops. As shown in
AM
Fig. 3.4, the four-wire transmit and receive sides of the hybrid
feed into two of its ports. The two-wire cable pair and a bal-
ancing network feed into the other two ports. If the balancing
TE
Echo Path
4-wire Receive
Pair
path, and where they are identical the echo signal is canceled
out. Echo cancelers are built into the IXC’s network, and are
not needed in the local exchange.
T carrier is excellent in the trunking network where
crosstalk can be controlled by use of dual cables or shielded
cables. It is not an ideal medium for the subscriber loop, how-
ever, because of crosstalk limitations. The line-coding method
of 1 bit per symbol change is not as efficient as newer line-
coding methods that are used in ISDN and other services.
Therefore, T carrier is generally replaced in the local loop by
one of several DSL technologies, which are the subject of the
next chapter.
4
Digital Subscriber Line Access Technology
Copyright © 2002 by The McGraw-Hill Companies, Inc. Click here for Terms of Use.
50 Access Technologies: DSL and Cable
Local Loop
PC
Bay Networks
51
DSLAM
ISP
LEC Central Office
Telephone
DSL Standards
One problem with DSL has been a glut of standards. The serv-
ice comes in an alphabet soup of options going by the names
of HDSL, SDSL, IDSL, ADSL, RADSL, and VDSL, often
collectively described by the term xDSL. Each of these has
different characteristics and fits a different niche in the mar-
ketplace. To add to the confusion, two different modulation
methods—discrete multitone (DMT) and carrierless ampli-
tude phase (CAP)—are used. The muddle isn’t quite as severe
as it seems at first. High-speed DSL (HDSL), and to some
degree single-pair DSL (SDSL), are used by the carriers them-
selves to provision point-to-point T1/E1, so the subscriber is
insulated from these. The industry itself drops the first letter in
its promotions, so users don’t have to know what variety of
DSL they are getting. Table 4.1 lists the types of DSL and their
approximate transmission ranges.
The inability of ILECs to respond to service requests has
hampered DSL’s growth. Many users have complained about
long delays in receiving services they have ordered and then
difficulties in getting the service to work. Technicians must
sometimes make multiple trips to get the PC and modem prop-
erly configured. To resolve these issues and to bring some
order to the chaos of multiple standards, the industry formed
the Universal ADSL Working Group in 1998 to develop a sin-
gle standard. The carriers’ principal objective was to make it
easy for subscribers to purchase self-configuring modems and
Table 4.1 DSL Types
Type of Upstream Downstream Range in
DSL Acronym Bandwidth Bandwidth ft (m)
Asymmetric ADSL 16–640 kbps 1.5–9 Mbps 18,000 (5,500)
53
High-bit-rate HDSL 1.544 or 2.048 Mbps 1.544 or 2.048 Mbps 12,000 (4,000)
ISDN IDSL 144 kbps 144 kbps 18,000 (5,500)
Single-pair SDSL 1.544 or 2.048 Mbps 1.544 or 2.048 Mbps 12,000 (4,000)
Splitterless G.lite 16–640 kbps 1.5–6 Mbps 18,000 (5,500)
Very-high-bit-rate VDSL 1.5–2.3 Mbps 13–53 Mbps 1,000–4,500 (330–1,500)
54 Access Technologies: DSL and Cable
DSL Technology
DSL lives in an environment that was originally designed for
voice transmissions. As discussed in Chap. 3, the principal
limiting factor is crosstalk. Since cable plant is called on to
carry high-speed signals that were not intended for copper
wire, care must be used to prevent interference from various
types of line signals. DSL can be categorized by two primary
criteria: the line coding and modulation methods.
Coding Methods
HDSL and ISDN DSL (IDSL) use the entire bandwidth of
the cable pair by applying a DC signal directly to the cable
pair. These types of DSL cannot coexist with a POTS line.
Other types of DSL separate voice from data with filters
and do permit the use of a POTS line. High-speed modems
encode information in symbols instead of a raw bit stream.
By using complex coding schemes, a single symbol can be
made to represent some number of consecutive bits.
As we have discussed, T1/E1 uses a simple coding scheme
that encodes only 1 bit per symbol. ISDN improves on that
by using 2B1Q line coding. Each pair of bits represents one
quaternary signal. Higher modem speeds are achieved with
QAM. With QAM, two carriers, each having the same fre-
quency, are phase-shifted 90˚ with respect each other. One
signal is called the I signal and the other the Q. Each carrier
is amplitude-modulated with half the data. The two signals
are combined at the source and transmitted to the receiver,
where the signals are separated and demodulated to produce
the original data stream.
Early QAM modem systems used eight phases and two
amplitude levels to transmit 4 bits over each symbol. These
56 Access Technologies: DSL and Cable
Modulation Methods
The second issue is the modulation method. DSL uses either
DMT or CAP. DMT separates the spectrum above the voice
Y
band into 256 narrow channels called bins. Each bin is 4 kHz
FL
wide. Data is modulated onto each channel using QAM modu-
lation with a 4-kHz symbol rate, resulting in up to 60 kbps per
AM
bin. A simple splitter separates voice from the data signal. The
modems detect which channels may be impaired and spread the
data to unimpaired channels. The upstream direction can use
bins 6 to 38, which is about 25 to 163 kHz. Bins 33 to 255 (142
TE
Spectral Compatibility
The degree to which DSL signals can coexist in the same
cable is a matter of concern to the ILECs. The limiting fac-
tor is usually near-end crosstalk. NEXT can be controlled by
limiting the nonvoice signals that are assigned to particular
groups of pairs. The NEXT problem is most severe nearest
the central office because the cable density is highest there.
DSL systems therefore use the lowest frequencies for the
upstream direction to compensate for the fact that the higher
the frequency, the greater the coupling from one pair to
another. To limit interference, each DSL standard includes a
power spectral density (PSD) mask that describes the PSD by
frequency in the upstream and downstream directions.
The degree of crosstalk that one DSL type imposes on
another defines their spectral compatibility. The degree of
compatibility depends on such issues as the spectrum the
DSL type uses and its placement in the cable.
Types of DSL
Each of the DSL types listed in Table 4.1 has a particular
function. This is usually not a matter of concern to the sub-
scriber because the service provider determines which type
will be used. This section briefly discusses the characteristics
and applications of each type.
HDSL
T1/E1, which is designed as a trunk carrier, doesn’t fit well
into exchange cables. Not only is repeater spacing critical, but
also, crosstalk considerations limit the number of systems that
can be installed in a 50-pair cable complement. HDSL is
designed to overcome these limitations. The ILECs have used
58 Access Technologies: DSL and Cable
SDSL
SDSL (also known as symmetrical DSL) serves the same
market as HDSL, with two important functional differences.
Instead of dual cable pairs, it provides T1/E1 service on a sin-
gle cable pair and it also derives a POTS line under the data
signal. It transmits and receives in the same band of frequen-
cies using an echo-canceling protocol, making it susceptible
to NEXT. SDSL uses 2B1Q line coding that is adopted from
basic rate ISDN.
IDSL
IDSL provides 128 kbps of bandwidth using the same proto-
col as BRI. The main difference is that IDSL terminates in a
router instead of an ISDN switch port. This precludes it from
carrying a voice signal. This technique has limited application
because its limited bandwidth cannot compete effectively
with other DSL technologies.
VDSL
VDSL was developed as a means of providing video on
demand in fiber-to-the-curb implementations. Fiber is used to
bring the video signal to an access node with twisted-pair wire
supporting the last span. VDSL supports data rates as high as
52 Mbps, which is enough to carry a DS3 over short spans of
copper wire. Rates of 52 Mbps can be supported over a 1000-
ft (330-m) range, but the rate drops to about 13 Mbps beyond
4500 ft (1500 m). The technology is asymmetric, with upstream
speeds of 1.5 to 2.3 Mbps. It can be overlaid over POTS lines.
ADSL
Asymmetric DSL (ADSL) can carry as much as 1.5 Mbps
downstream over a range of up to 18,000 ft (5,500 m). Data
Digital Subscriber Line Access Technology 59
DSLAM
The DSLAM separates voice-frequency signals from data at
the CO end. The DSLAM aggregates data traffic from several
loops into the backbone. In the case of ILEC-provided DSL,
the backbone is usually ATM. Competitive access providers
may use IP. ISPs and other network providers connect to the
network with PVCs.
A multiservice DSLAM can handle traffic from SDSL,
IDSL, and ADSL. It may also support video access in addi-
tion to Internet access. Multiservice capability may also
include multiple service types such as frame relay, ATM,
TDM, and IP running to the subscriber from a single plat-
form. Frame relay is provisioned to the subscriber using
TDM or ATM.
VoDSL
DSL started out to keep voice and data separated by means
of splitters and filters, but several new protocols and prod-
ucts have been developed or are under development to merge
voice and data at the circuit level. The products compress
the voice signal down to a fraction of the 64-kbps bandwidth
of a conventional voice channel. Voice can be compressed to
8 kbps with only a small loss of intelligibility. The quality is
equivalent to a solid cell phone connection. The protocols
add some overhead, so the number of channels is less than a
straight multiple of the compressor output, but the voice-
carrying capacity is substantial. Note, however, that faxes
and modems require an uncompressed channel.
The amount of bandwidth the voice channel occupies
depends on the protocol used. ATM is the most common pro-
tocol. With 5 bytes of header per cell, the overhead is approx-
imately 10 percent. Equipment vendors claim from 16 to 20
voice channels per DSL line, but of course that depends
on the actual data rate the line can sustain and also on the
amount of bandwidth needed for data.
5
Cable Access Technology
Copyright © 2002 by The McGraw-Hill Companies, Inc. Click here for Terms of Use.
62 Access Technologies: DSL and Cable
Two-Way Systems
Two-way cable systems are constructed on the same princi-
ples as one-way systems, but with bidirectional amplifiers.
Filters split the signal into the high band for downstream
transmission and the low band for upstream transmission.
The ratio of downstream to upstream bandwidth is heavily
weighted in the downstream direction. The 5- to 40-MHz
range is used for upstream bandwidth, with a guard band
about 15 MHz wide to separate the two directions of trans-
mission. The upstream direction shares frequencies with
66 Access Technologies: DSL and Cable
just at the time other stations have data to send, the stations
wait until the network is idle.
The Aloha situation is similar to that in a cable network.
Stations can hear the transmissions of upstream stations, but
they cannot hear downstream stations. If stations simply lis-
ten to the channel and wait for an idle moment to transmit,
the upstream stations grab a disproportionate share of the
bandwidth and squeeze out the downstream stations. This
can be resolved by using a protocol known as slotted Aloha.
Stations are granted a time slot within which they can trans-
mit. The headend is responsible for ensuring that bandwidth
is evenly distributed by assigning transmission timeslots to
active stations.
Proprietary methods were used for years on cable net-
works, but the result was that users could obtain only the
types of modems that the cable operator certified for its net-
work. Proprietary equipment tends to be more expensive than
standard equipment, and also limits the growth of the tech-
nology, which led the cable industry to develop a protocol
that all could employ.
DOCSIS
DOCSIS was developed by CableLabs, Inc., a consortium of
equipment manufacturers that collaborated on creating a
standard for data transmission over cable. In Europe, DOC-
SIS is known as a Euro-DOCSIS, and is derived from the
U.S. version. The European cable community adopted the
standard in early 2000. DOCSIS consists of several compo-
nents. The cable modem (CM) connects the subscriber’s PC
to the cable network. At the headend is the cable modem ter-
mination system (CMTS) and a variety of specialized
servers. This equipment operates with the subscriber’s
modem as either a bridge or router. In between is the HFC
plant, which forms the radio frequency (RF) link between the
CMTS and the CM. DOCSIS specifies the RF physical layer
with respect to modulation methods and symbol rates. It
specifies modem initialization procedures, security, and data
70 Access Technologies: DSL and Cable
188 bytes are MPEG header and information and the remain-
der are parity bits to correct errors. The purpose of forward
error correction is to reduce the number of frames that must
be retransmitted because errors occurred. Each frame starts
with a packet identifier (PID). Standard PIDs are assigned
for video, audio, clock, and other data such as the program
guide. MPEG-2 can multiplex multiple audio and video pro-
grams together on the same bit stream.
One feature of DOCSIS is the provision of various class-
es of service. Cable providers can offer priority to customers
who are willing to pay, just as the airlines offer first-class,
business class, and tourist seats. Although some compromis-
es are required to use cable as an access medium, it is a reli-
able system that provides sufficient bandwidth to meet near-
ly all residential requirements. From the cable operator’s
standpoint, it is a profitable service. The main problem is that
as demand grows, response time slows. The only solution is
to segment the coaxial portion of the network, which is an
expensive proposition and not always practical.
6
Wireless Access Technology
Copyright © 2002 by The McGraw-Hill Companies, Inc. Click here for Terms of Use.
76 Access Technologies: DSL and Cable
Cellular Dial-up
Within limits, cellular dial-up can be an effective access
option. The first limitation is low data throughput, which
makes it unsatisfactory for Web surfing. Ordinary commer-
cial modems may not work well with cellular, particularly at
high speeds. The modulation methods of these modems are
complex, and they do not gracefully handle interference,
fades, and signal dropout. Handoff between cells causes a
momentary interruption, and may cause the connection to
drop. Furthermore, many modems are designed to operate
only after they recognize dial tone, which cellular does not
provide. Therefore, special cellular modems that can adjust
speed to the signal conditions are required.
For occasional use such as mobile file transfer, cellular is a
satisfactory alternative, but it is expensive for short messages.
Most cellular operators levy a one-minute per-call minimum
charge, which makes cellular a poor choice for short transac-
tions that may last only a few seconds. Also, the setup time is
long compared to other alternatives. Cellular has the advan-
tage of good coverage. In general, its applications are similar
to those of the PSTN: it is acceptable for facsimile and file
transfers, but poor for short, bursty messages.
Next-Generation Cellular
The cellular industry is touting so-called third-generation
cellular as the solution for Web surfing from a hand-held
mobile device. The first cellular generation was analog fre-
quency-division cellular, some of which is still in operation.
As the industry ran out of capacity in the cellular bands,
which are between 824 and 894 MHz, second-generation
digital cellular was introduced to make more efficient use of
the spectrum. Most of the world elected to use the global sys-
tem for mobile communications (GSM) modulation method.
In the United States, the FCC elected to let the market deter-
mine the most effective method. As a result, two methods are
used, neither of which is compatible with GSM. AT&T uses
Wireless Access Technology 81
for no other reason than the fact that it occupies the same fre-
quencies as 802.11b wireless and that coexistence may be
a problem.
CDPD
CDPD is a good alternative for obtaining wireless coverage in
metropolitan areas and along major highways because it offers
the same coverage area as cellular if the service provider has
elected to equip its cell sites for the service. CDPD is a packet-
switched data service that rides on top of cellular and uses idle
analog channel time. It can be added to existing cell sites at a
moderate cost. Carriers charge by the packet or kilobyte
instead of by connect time, and the long call setup time and
minimum connect time charges are eliminated. This makes
CDPD good for short, bursty messages such as point of sale,
dispatch, package tracking, telemetry, and e-mail. CDPD is
available in most metropolitan areas.
CDPD operates at 19.2 kbps using a TCP/IP type of pro-
tocol, which raises the problem of IP addressing because the
subnet is mobile. For CDPD to be entirely effective for some
applications, a laptop user should be able to disconnect from
the LAN, travel to another location while remaining in con-
tact with the network through a wireless connection, recon-
nect to the LAN at the distant location, and become part of
the network again. The process is possible today, but the user
needs to understand how to do it. Therefore, it is not yet fea-
sible for the true mobile laptop or PDA application.
To implement CDPD, carriers install mobile database
stations (MDBSs), which retrieve packets from the wireless
network, and a mobile data intermediate system (MDIS),
which routes them. Frames are picked up by the MDBS and
handed off to the MDIS. Mobile stations use a protocol called
Digital Sense Multiple Access with Collision Detection
(DSMA/CD) for access to the network. The access method is
similar to Ethernet’s Carrier Sense Multiple Access with
Collision Detection (CSMA/CD). A station wishing to trans-
mit listens to the outbound channel to determine if a carrier is
Wireless Access Technology 83
LMDS
In 1998 the FCC auctioned 1.3 GHz of spectrum in the 28-
to 31-GHz range for LMDS. In each geographical area, WLL
providers bid for the right to use frequencies known as the A
block, with 1150 MHz of bandwidth, and the B block, with
150 MHz of bandwidth. LMDS is intended for such services
as multichannel video, interactive gaming, streaming video,
telephone service, and Internet access. LMDS is a point-to-
multipoint service. The service provider locates a hub in the
center of a serving area that may serve several thousand
homes. Subscribers are equipped with small rooftop anten-
nas and transceivers and feed data into the pipeline using an
ATM-like protocol or IP. Multiple hubs are linked with fiber
optics. The bandwidths available to customers range from
one DS-1 to as much as OC-3. Figure 6.1 shows the archi-
tecture of a typical LMDS network. In Canada a similar serv-
ice is called local multipoint communication systems, and
similar services are available in many European countries.
LMDS is intended to allow service providers to bypass the
copper local loop with an economical service that can be
deployed rapidly. There is no need to dig up the streets to
place cable, and the scalable nature of the service enables
providers to meet customer demand in a few days after the
hub is installed. A major objective is to defer the amount of
unused investment. With copper wire plant, the ILECs must
build initially for what they forecast will be the ultimate serv-
ice demand because of the high cost of adding more. This
Y
PSTN
FL
86
AM
OC3 to 12
MMDS
MMDS was originally intended as a wireless replacement
for cable. For TV signals with 6 MHz of bandwidth, 33
channels can fit into the 200 MHz of spectrum that is allo-
cated for MMDS use between 2.5 and 2.7 GHz. The archi-
tecture of MMDS is similar to that of LMDS, consisting of
headend equipment similar to that used in a CATV system
and receiving equipment at subscriber locations. For video
reception, a set-top converter demodulates the incoming sig-
nal to the frequency of a conventional television channel.
The MMDS signal is transmitted from an omnidirectional
antenna. Repeaters may be used to extend the range or to fill
dead areas caused by shadows in the coverage area.
Although MMDS was initially intended for one-way
video, it is now authorized for two-way service, making it
applicable to Internet access as well. Typical service offer-
ings provide downstream transmission rates of 1 Mbps or
higher, scalable up to 10 Mbps, and upstream speeds up to
512 kbps, which makes MMDS competitive with DSL and
cable access. The MMDS spectrum is shared with instruc-
tional television fixed service (ITFS), which is intended for
distance-learning video. The 6-MHz video channels can be
modulated with data signals using the same concepts, and in
many cases the same hardware, as cable modems.
Satellite Service
One of the major attractions of satellite service has always
been the elimination of the local loop. Satellite is ideal for
broadcast applications where it is desirable to uplink a signal
from one location and downlink it to a vast area. This is the
nature of the direct broadcast satellite services that have been
competing with cable for the past few years. These services
are acceptable for Internet use as well, except that their one-
way nature means that a telephone return is required, with its
attendant disadvantages.
Wireless Access Technology 89
I f you could see a few decades into the future, what might
the access network look like? No one is certain at this
point because there are so many conflicting opinions. It
seems clear, however, that the twisted-pair wire and coaxial
cable of the past must give way to a medium with greater
bandwidth. Based on what is now known, fiber optics will be
that medium. Today, copper wire, coaxial, and wireless alter-
natives all rely on fiber trunking to bring information to
neighborhood nodes. There, depending on the service
provider, the information jumps on a short twisted-pair wire
run, coax, or a wireless link of some kind. The fiber stops at
the neighborhood node because carrying it all the way to the
subscriber is too costly.
As long as the infrastructure can be installed before streets
and sidewalks are installed, the cost of fiber is supportable,
but fiber still has to fit into the existing service provider’s
91
Copyright © 2002 by The McGraw-Hill Companies, Inc. Click here for Terms of Use.
92 Access Technologies: DSL and Cable
The PON
The PON is a logical physical structure for single-pipe infor-
mation services in the local loop. Conceptually, the service is
not much different than cable except for the physical medi-
um, which provides significantly more bandwidth. Figure 7.1
shows the PON architecture. It is typically installed with a
96 Access Technologies: DSL and Cable
ONU
ONU
Optical ONU
Feeder Passive
Line ONU
Fibers Splitters
Termination ONU
ONU
Central Office
ONU = Optical
Network Unit
ONU
ONU
Copyright © 2002 by The McGraw-Hill Companies, Inc. Click here for Terms of Use.
100 Access Technologies: DSL and Cable
T1/E1 Access
If a full point-to-point T1/E1 can be justified to a branch office,
the same bandwidth is available in both the access circuit and
the backbone and the facility has plenty of stability to handle
both voice and data. T1/E1 is an ideal medium if the bandwidth
requirements justify the cost. This is often the case with branch
offices, but it is rarely justified for telecommuting.
DSL Access
If the branch or home office is within range, DSL can be an
ideal medium. Scalability problems that are typical of T1/E1
service are resolved, and the service provider’s architecture
104 Access Technologies: DSL and Cable
may solve the QoS issue. As we have seen, most ILECs use
ATM in the access circuit and in their backbone. If a single
carrier provides the service end-to-end, the quality should be
controllable within tight limits and could be enough to sup-
port both voice and data. Many PBX manufacturers provide
a proprietary IAD that works with their digital telephone.
Security is tight, and the telecommuter or branch office
worker should have facilities equivalent to a station collocat-
ed with the PBX.
Quality is less predictable when more than one service
provider is involved. In a metropolitan area with two ILECs,
if each one has an ATM network with a network-to-network
interface, the service may be equivalent to the ATM network
of a single provider. When the network extends between two
metropolitan areas, it may be necessary to connect them over
an IXC’s IP or ATM backbone. Some IXCs provide service
level agreements (SLAs) that may be high enough in quality
to support a branch office or telecommuting application. Be
aware, however, that many IXCs quote their SLAs as an aver-
age over 30 days. To obtain satisfactory quality of service,
you must know the worst-case delay and jitter figures to
determine whether quality will be satisfactory to support the
application.
Dial-up access
For some access applications, dial-up is the only feasible
alternative. Despite the bandwidth limitations of the PSTN, it
is almost universally available, although not necessarily at a
low cost. Many ISPs provide dial access to the Internet as
well as e-mail. A premium price may be charged, and access
is available almost anywhere, but it is almost precluded as a
telecommuting medium because of lack of bandwidth. Some
PBX manufacturers provide devices to enable the use of a
digital telephone over an analog loop, but these are not high-
ly effective for full-time telecommuting.
If bandwidth limitations prevent the use of dial-up, then
ISDN may be an alternative. Basic rate ISDN suffers from
the same problems as DSL, however, in that it has an 18,000-
ft (5,500 m) limitation. Furthermore, unless the channels are
bonded, the bandwidth may not be enough of an improve-
ment over analog to justify the extra cost. BRI is often
metered, even in locations where flat-rate analog is available,
and an always-on connection to the Internet or headquarters
can be prohibitively expensive. Finally, BRI suffers from a
general lack of availability in the United States. Many
European countries have much better ISDN availability, but
many central offices in the United States have not been
equipped for BRI. Some multioffice wire centers have ISDN
available in one switching system, but a number change may
be required to switch from analog to ISDN service.
One promising service, always-on dynamic ISDN
(AO/DI), uses the D channel for a full-time connection to the
Internet. If more bandwidth is needed for Web surfing or to
transfer an e-mail file attachment, the service brings a B
channel online long enough to provide the necessary band-
width, after which it disconnects. The theory of AO/DI is
106 Access Technologies: DSL and Cable
109
Copyright © 2002 by The McGraw-Hill Companies, Inc. Click here for Terms of Use.
110 Access Technologies: DSL and Cable