UNIT IV

DIGITAL SUBSCRIBER ACCESS

INTEGRATED SERVICES DIGITAL NETWORK11.5 HYBRID FIBER COAX SYSTEMS 1

In addition to the digitization of the internal portions of public telephone networks, a lesser
known but also significant change involved the development of common-channel signaling (CCS)
for network control. Both the digitization and the use of CCS started at internal portions of the
network and migrated toward the periphery. Except for some special data service offerings and a
few network-based features derived from the signaling network, these facilities provided no direct
benefit to the end users. As shown in Figure, ISDN is a service offering that extends access to both
of these facilities to the end user. Access to the digital transport facilities occurs on 64-kbps bearer
(B) channels while access to the signaling network occurs on 16- or 64-kbps signaling (D) channels.
Major features or benefits made available by these channels are listed in Tables.

Two levels of digital access to the ISDN network have been standardized: basic rate access
and primary rate access. As shown in Figure, the (worldwide) basic rate interface (BRI) standard is
also referred to as a 2B + D interface. In North America, the primary rate interface (PRI) standard is
sometimes referred to as 23B + D while the ITU-T counterpart is 30B + D. The North American
PRI is fundamentally a 1.544Mbps DS1 signal with the D channel replacing one of the 24 message
channels (usually the last one). To achieve a 64-kbps clear channel capability, a B8ZS line code is
used to eliminate one’s density requirements and common-channel signaling frees up the signaling
bits so the full 64-kbps bandwidth is available for user data. The ITU-T

Figure Integrated Services Digital Network access to circuits, channels, leased lines, and
common-channel signaling.

1
Features of ISDN B Channels
 End-to-end four-wire digital circuits: no loss or echoes for voice circuits using digital
instruments
 Shared 2network
DIGITAL SUBSCRIBER ACCESS
access for voice, data, and leased lines
 Relatively high bandwidth data channels (64 kbps)
 Lower error rates than typical voiceband modems
 In-service performance monitoring
 Possible expansion of speech bandwidth because elimination of tandem encodings allows
greater one-time quantization errors
 PRI is a 2.048-Mbps El digital signal with the D channel occupying the signaling channel
(time slot 16). Because a single D channel can support more than one PRI, 24B and 3IB
interfaces are allowed for additional PRIs in a group of PRIs.

ISDN Basic Rate Access Architecture

An ISDN basic rate access line is a standard copper pair that has been specially conditioned
to support a bidirectional 160-kbps aggregate data rate. Transmission technology required for basic
rate access is generally referred to as the digital subscriber loop (DSL). Complications arise when
using existing analog pairs. The principal considerations are bridged taps and wire gauge changes,
both of which cause reflections that impact higher speed digital signals. To allow flexibility in the
selection and deployment of the DSL, the ITU-T basic rate specification does not define a two-wire
transmission standard. Instead, it establishes an interface standard that assumes the presence of a
network termination module that converts any chosen transmission system to the standard interface.
In the interest of supporting deregulated customer premises equipment, the Exchange Carriers
Standard Association in the United States established a basic rate transmission standard so CPE
equipment could connect directly to the transmission link or select network termination modules
from alternate vendors. Figure depicts the architecture and associated terminology of a North
American BRI.
Features of ISDN D Channels
 Signaling simultaneous with active connections
 Calling number identification
 Far-end supervision
 User-to-user message transfer
 Telemetry for fire alarms, security, meter reading, etc,

19
 Access to packet-switching network
 Support for network services such as multiple directory numbers sharing one or more B
channels, trunk group blocking statistics, and identification of calling number for abandoned or
11.6 VOICEBAND MODEMS 3

blocked calls
Basic rate and primary rate access to ISDN.
Module Definitions
NT1: A network termination module for layer 1 functions that provides physical and electrical
termination of the transmission link only. In essence, the NT1 isolates the user from the
transmission technology but does not demultiplex or process D channel messages.
NT2: A second level of network termination that implements functions associated with layers 2 and
3 of the OSI protocol stack. Thus, NT2 equipment extracts and processes D channel messages.
Representative NT2 equipment includes PBXs, multiplexers, or LAN gateways.
TE1: Type 1 terminal equipment such as a digital telephone that complies with the ISDN S
interface recommendation.
TA: Terminal adapter used to convert from an arbitrary (R) interface to the ISDN S interface.
TE2: A non-ISDN terminal that requires a terminal adapter to interface to the ISDN S interface.
Prevalent examples of a TE2 equipment are analog telephones or asynchronous (RS-232) data
terminals.
Reference Points
U: Interface to the two-wire transmission line.

T: CCITT ISDN interface defined in Recommendation 1.430.

3
 S: Interface to NT2 equipment identical to a T interface.
R: A non-ISDN interface such as an analog tip and ring.
S/T Interface
The S/T4 DIGITAL SUBSCRIBER ACCESS
interface is defined in ITU-T recommendation 1.430 to be supplied by network
termination equipment (NT2/NT1). It is intended for customer premises installations only. (No
overvoltage protection is prescribed.) The most significant aspects of the S/T interface are:
 Four-wire facility (one pair for each direction)
 One kilometer maximum required distance
 Alternate space inversion line code (which is the inverse of an AMI line code: see Figure 11.4)
 Point-to-point or point-to-multipoint configurations
 Data rate of 192 kbps with 48 kbps of framing, control, and synchronization
The frame structure at reference points S and T is shown in Figure 11.5. As indicated, each
250-msec frame contains 48 bits. Thirty-eight of these bits are common to both directions of
transmission and are defined as follows:
16 bits in first B channel (Bl)
16 bits in second B channel (B2)
4 bits in the D channel 1 bit in the framing channel F
1 bit in the auxiliary framing channel FA
The remaining 10 bits are assigned different functions depending on the direction of
transmission. From the TE to the NT all remaining 10 bits are defined as L bits, which are used to
maintain dc balance. The definition of the 10 bits from the NT to the TE
2 L bits for maintaining dc balance

Alternate space inversion code

S/T frame
structure.

4 E bits
that echo D bits received from the TE(s)

19
 1 A bit for activation
1 N bit, which is the complement of the FA bit 1 M bit for multiframe
identification
1 S bit for S channel 11.6 VOICEBAND MODEMS 5

Figure indicates that the framing bit F is always a binary 0. Even though a positive voltage
level is indicated, either a positive or negative voltage is allowed so the receivers are not sensitive
to wiring polarity. A transmitter always produces the same level, however, so the receiver always
receives the same polarity in every framing bit. As an aid in rapid acquisition of the framing
pattern, the framing bit always represents a line code violation (it is the same polarity as the
previous 0). To maintain dc balance, an L bit with the opposite polarity of the F bit always follows
the F bit. The first 0 in a data block following the L bit is encoded with the same polarity as the L
bit, which implies another line code violation. Direct-current balancing of this violation is the
purpose of the L bit at the end of each data block, which also assures that the next (fixed-polarity) F
bit produces a line code violation.
The reason for the additional L bits in the frame from the TE to the NT arises because more
than one TE can be connected to the S interface as a passive bus. Because the TEs transmit
independently of each other, each individual transmission (D channel bits and B channel bytes) is
individually dc balanced.
Passive bus operations are also the reason for the existence of the NT-to-TE E bits. Multiple
station access to the D channel is controlled by having a terminal wait for an idle code on the NT-
to-TE D channel before transmitting on the TE-to-NT D channel. When a terminal begins D
channel transmission, it monitors the incoming E bits. If an incoming E bit does not match the
previously transmitted D bit, that terminal stops transmitting and waits to seize the channel at a later
time. Two levels of priority are defined for accessing the D channel. Signaling information is the
highest priority while user packet messages are the lower priority. All terminals on a passive bus
have equal priorities within each level.
S-bus connections

ISDN U Interface
Prior to the establishment of a standard U
interface by the Exchange Carriers Standards
Association in North America a number of basic rate transmission systems were developed by

5
telephone equipment suppliers around the world and put into service.
One of these involves the use of time compression multiplex (TCM). TCM, developed by
NEC in Japan, provides full-duplex transmission on a single pair of wires by alternately transmitting
bursts of data6 DIGITAL
in each direction. For this reason it is sometimes referred to as “ping-pong”
SUBSCRIBER ACCESS

transmission. One big advantage of TCM transmission is that near-end crosstalk is avoided because
a station is never receiving while transmitting. The big disadvantage is that the burst data rate must
be more than twice the desired data throughput.
AT&T in the United States also developed a BRI transmission system for the No. 5 ESS end
office switching system available with generic releases 5E4 and 5E5, this system has a 160-kbps
data rate utilizing a 50% duty cycle AMI line code. Full-duplex transmission is achieved by
simultaneous transmission in both directions using hybrids and echo cancelers (ECs) to separate the
two signals. Beginning with generic 5E6 the No. 5 ESS supports both the AT&T (Lucent) U
interface (referred to as a 5E4/5E5 U interface) and the ANSI U interface.
Like the 5E4/5E5 U interface, the ANSI U interface uses simultaneous transmission in both
directions with echo cancelers and a data rate of 160 kbps. The major difference is the use of a four-
level line code referred to as 2B1Q (two binary digits in one quaternary digit). Thus, the symbol
(baud) rate on the line is 80,000 symbols/sec. Because the line code itself does not prevent dc
wander, dc restoration is necessary. A 2B1Q line code was chosen primarily because the lower
symbol rate minimizes the two dominant transmission limitations in this application: intersymbol
interference and near-end crosstalk [6].
The frame format and superframe structure of the ANSI U interface are shown in Figure
11.8. Each frame consists of 240 bits containing 18 framing bits, 216 payload bits (12 fields of 18
2B + D data bits), and 6 overhead bits. Because the frame rate is 667 frames/sec the data rate is 160
kbps. The 6 overhead bits are organized as a block of 48 bits in an eight-frame superframe.
Functions included in the overhead bits are 24 bits of embedded operations channel, 1 activation bit,
1 deactivation bit, 1 far-end block error bit, 12 CRC bits, and 9 fixed 1 bits. All bits except the
framing bits are scrambled for transmission.

19
 ISDN DSL TX/RX block diagram

11.6 VOICEBAND MODEMS 7

ANSI U interface frame and superframe structure.

ISDN D Channel Protocol

The D channel protocol is defined in two separate series of ITU-T recommendations: the I
series and the Q series. The data link layer (LAPD) is defined in 1.441 or Q.921. This protocol is
similar to LAPB of the X.25 standard except Q.921 allows more than one logical link. (Therefore
separate “connections” can exist for signaling, packet network, or far-end terminals.) The main
functions of the data link layer are message sequencing, error checking and retransmission, and data
layer link recovery. The network layer of the D channel protocol is defined in 1.451 or Q.931. This
layer provides connection setup, alerting, routing, and release of ISDN calls. When a B channel
accesses a packet network, the X.25 protocol is used while in the connected state.

HIGH-DATA-RATE DIGITAL SUBSCRIBER LOOPS

A basic rate ISDN digital subscriber loop provides an aggregate, bidirectional data rate of
160 kbps on a single pair of wires. This section describes several transmission techniques that allow
much larger bandwidths on copper wire pairs. These new techniques are enabled by the availability
of low-cost, high-performance digital signal processing. Several versions of high-speed digital
subscriber lines have been developed. The various versions are collectively referred to as xDSL
Asymmetric Digital Subscriber Line
ADSL allows for high data rates to the subscriber and moderate to low data rates from the
subscriber to the network. ADSL technology was originally conceived as a means of delivering
switched digital video services over a copper loop, which obviously do not require high data rates
from the subscriber. Although video applications did not

7
Versions of Digital Subscriber Lines

DSL Digital Subscriber Line (ISDN basic rate)

ADSL Asymmetric DSL (9 Mbps downstream, 640 kbps upstream)8
8 DIGITAL SUBSCRIBER ACCESS
HDSL High-bit-rate DSL (T1/E1 service on two pairs)
SDSL Single-line DSL (T1/E1 service on one pair)
VDSL Very high bit rate DSL (52 Mbps downstream, 2.3 Mbps upstream)3

materialize, the asymmetric data rate is also suited for Internet access. The philosophy behind the
asymmetric data rates is that subscribers typically need to receive high- bandwidth data (for
Internet file downloads) but normally need to transmit (query) at a relatively low data rate. The
subscriber’s receive data rate on ADSL varies between 1.5 and 9 Mbps while the subscriber
transmit rate varies between 16 and 640 kbps/ The specific data rate utilized depends on the
transmission quality of the particular wire pair.
In contrast to ISDN channels, which represent extensions of the digital facilities of the
public switched telephone network (PSTN), ADSL channels are separated from the public network
at the line interface of the PSTN. As shown in Figure 11.9, ADSL lines terminate at a telephone
company central office (or remote terminal) where the data streams are forwarded to and received
from a facility that is separate from the telephone network. Transmission between the line interface
and an Internet service provider, for example, is typically provided with an ATM connection.
ADSL has two major advantages over ISDN access. First, ADSL data rates provided to the
subscriber are significantly higher than the 128-kbps ISDN basic rate. Second, ADSL piggy backs
digital transmission on a standard analog telephone wire pair. Thus, existing analog telephones are
retained on ADSL but are either replaced by digital phones or are connected through conversion
devices when ISDN is utilized.
Two versions of ADSL transmission links have been developed: carrierless amplitude and
phase (CAP) modulation and discrete multitone (DMT) modulation. CAP is the first version
deployed but DMT has been selected as the standard [8], Because DMT makes more intensive use
of DSP, a DMT implementation typically requires more power—a significant consideration for
remote terminal deployment. CAP, on the other hand, its not generally considered to be as flexible
as DMT in achieving maximum data rates on some wire pairs or in some interference
environments. For a thorough comparison of the two alternatives
DMT Implementation
Basic parameters of the standard ADSL DMT implementation are provided

19
A less ambitious version of ADSL referred to as G.Lite only attempts to achieve 1.544 Mbps
downstream and 384 kbps
upstream.
ADSL network configuration.
11.6 VOICEBAND MODEMS 9

ADSL DMT Implementation Parameter

a modulator and an FFT as the corresponding demodulator. The following paragraphs identify the
basic function of each block.
PRS Source: A pseudorandom sequence generator provides a prescribed sequence for
characterizing the channel during a training period. Characteristics of the channel that are
determined during the training period are attenuation and phase distortion across the band,
noise/interference levels across the band, and the information capacity of individual
subchannels.
Channel Allocation: Provides assignment of data bits to individual subchannels according to the

9
 subchannel capacity determinations obtained during training.
FEC: A combination of Reed-Solomon and convolutional coding.
IFFT QAM Modulation: Conversion of data values to quadrature channel signal amplitudes
10 DIGITAL
and conversion to aSUBSCRIBER ACCESS
time-domain waveform using an inverse FFT.
D/A: Digital-to-analog conversion.
TX Filter: Bandpass filter to prevent interference into the voiceband and to smooth the discrete
DSP samples.
POTS Splitter: Used at both ends of the line to separate the analog voice from the data. (Simpler
versions of ADSL incorporate the splitter function in the modems to simplify installation.)
AGC: Automatic gain control to adjust overall receive level.
Slope/Delay Equalization: A front-end equalizer to partially flatten the frequency response and
equalize extreme delay variations in the channel,
A/D: Analog-to-digital conversion.

ADSL DMT block diagram

FFT QAM Demodulation: FFT conversion of time-sampled waveform to frequency domain where
data values are related to amplitudes of quadrature carriers.
Clock Recovery: A/D sample timing obtained by locking to pilot frequency. The desired sampling
rate is eight times the pilot frequency so l-of-8 phase ambiguity has to be determined by
monitoring framing/synchronization bit integrity.
Frequency-Domain Equalizer: Multiplication of complex (quadrature) frequency spectrum by
amplitude and phase equalization parameters obtained during training period.
Data Detection and Interleaving: Slicing of quadrature amplitude values to decode data and
subsequent generation of composite stream identical to original source data.
VDSL
VDSL is an expanded version of ADSL to achieve even higher bandwidths on particularly short

19
lines as would be available from remote terminals of fiber to the curb systems. A primary motivation
for VDSL is potential distribution of HDTV signals. Although several different modulation
techniques have been proposed for VDSL, a DMT version seems to be favored.
11.6 VOICEBAND MODEMS 11

DIGITAL LOOP CARRIER SYSTEMS

The primary purpose of a digital loop carrier (DLC) system is to reduce or eliminate copper
pairs from a central office to the vicinity of a group of subscribers. Even though the transmission
link from the central office to the DLC remote terminal is digital, the transmission links from the
remote terminal to the subscribers are typically conventional analog loops. Thus, the main purpose
of a DLC is not to provide digital subscriber access. Nevertheless, some DLC systems (particularly
fiber-based DLC systems) provide options for Tl, ISDN, or xDSL digital interfaces. Moreover, the
fact that the DLC remote terminal is relatively close to the subscriber locations facilitates the use of
these interfaces. (A short copper drop from the remote terminal has low attenuation and crosstalk
and is less likely to have loading coils, bridged taps, or multiple sections of wire with

Universal digital loop carrier system.

varying gauges.) Although a DLC is a natural mechanism to provide enhanced services through
ISDN or xDSL digital interfaces, mechanical packaging and power considerations of these interfaces
represent unique requirements with respect to strictly POTS applications.

Universal Digital Loop Carrier Systems

A universal digital loop carrier (UDLC) system can be interfaced to any switching system:
analog or digital. The interface between the local switching system (end office) of the public
network and the central office terminal (COT) involves individual circuits (e.g., individual analog tip
and ring connections). The multiplexed digital transmission links between the COT and the remote
terminal (RT) can be wire pairs or fiber. Each interface of a COT is paired with a corresponding
subscriber interface at the RT so the use of a UDLC is transparent to both the switch and the
subscriber. In its simplest mode of operation, the UDLC uses pure multiplexing between the COT
and the RT so that there is a one-to-one correspondence between a particular TDM channel and the
COT/RT interface pair. Some systems can also be configured with concentration wherein the

11
COT/RT pairs are dynamically assigned transmission channels. If the number of requested
conversations exceeds the number of channels, blocking occurs. The possibility of blocking
introduces non transparency and implies that some means of returning a reorder tone (fast busy) is
12 DIGITAL SUBSCRIBER ACCESS
needed in the RT.
UDLC installations are configured to match each particular interface of the central office
switch with a complimentary interface in the RT. A fully capable system must provide a wide
variety of interfaces such as loop-start line, ground-start PBX trunk, foreign exchange lines, and
coin telephone interfaces. In some early systems the configuration process involved nothing more
than physically installing matching interfaces in respective equipment slots of the COT and RT.
More recent systems typically utilize line units with multiple-service capabilities. These systems
can be configured electronically (i.e., no straps) with either a local or a remote management
interface.
Integrated Digital Loop Carrier Systems
Whenever a universal digital loop carrier system is interfaced with a digital switch, obvious
inefficiencies occur in terms of back-to-back demultiplexing-multiplexing and D/A-A/D
conversion. As shown in Figure, an integrated digital loop carrier (IDLC) system eliminates the
inefficiency by directly connecting the DLC TDM link to the digital matrix. Typically the direct
digital connections are either DS1 or El cross-connect signals. Thus, a fiber-based DLC system will
typically interface with the switch through multiplexing/demultiplexing equipment as some number
of digital cross-connect signals.

Integrated digital loop carrier system.

From a functional point of view an early IDLC systems was nothing more than a distributed
switching system wherein some line interfaces of the switch are moved to remote locations. Most
central office switch manufacturers provide such a capability— sometimes as simply as using
channel banks for analog line interfaces that can be colocated with the matrix or remoted. Central
office switch vendors also offer remote switching modules wherein some portion of the matrix itself
is remotely located. These configurations provide remote concentration for efficient use of the
connecting transmission link and, in some cases, provide local switching in the remote module so
connections between two ports of a remote module do not use the transmission link. (If remote local

19
switching is not provided, a connection between two ports of a remote module requires two channels
of the transmission link.)
Remote modules of a particular switching system vendor are often implemented with
proprietary signaling and management processes that preclude use of DLC
11.6 VOICEBAND equipment
MODEMS 13 from other
vendors (except when a DLC vendor develops the ability to emulate a particular switch vendor’s
remote modules). In the interest of opening IDLC applications to competing vendors, Bellcore
established an IDLC standard referred to as GR-303 that compliant switching system vendors must
support (possibly in addition to a proprietary IDLC capability). Included in the GR-303 specification
are definitions for signaling, provisioning, testing, alarm surveillance, and performance monitoring.
Due, in part, to a move to unbundle LEC local loop services GR-303 has assumed a much
broader scope than just an IDLC application. Because a GR-303 capability includes being able to
define and administer a myriad of switch interface types, the GR- 303 standard can be used for
interfacing other types of equipment such as xDSL equipment. ETSI has established a similar IDLC
standard for international (ITU) switching equipment referred to as a V5 interface.
Next-Generation Digital Loop Carrier Systems
The term next-generation digital loop carrier (NGDLC) has been adopted within the industry
to refer to DLC systems that adhere to GR-303 and provide additional configuration options and
interfaces. There is no precise definition of what constitutes an NGDLC system. Four basic
attributes are adherence to GR-303, optical fiber transmission capabilities (e.g., SONET/SDH),
generally larger line sizes, and an ability to interface with a central office operational support system
for diagnostics, alarms, and remote provisioning [14]. Other aspects commonly available in NGDLC
systems are depicted .

Next-generation digital loop carrier.

The most important aspect of an NGDLC system from a services point of view is the
availability of new services such as copper or fiber digital subscriber interfaces in addition to the
conventional POTS interfaces. The most desired digital interfaces are Tl/El, primary rate and basic
rate ISDN, ADSL, and VDSL. Other digital interfaces may also be provided for services like
broadband data and digital video. Although the diversity of interfaces belies use of a single multiple
service line interface for total electronic provisioning, the system must report inconsistencies

13
between installed hardware and the electronic database. A particularly desirable feature of the
system shown in Figure is drop-and-insert capability, which allows distributed access to a single
backbone digital route. GR-303 identifies star configurations, linear ADM distribution, and ADM
14 DIGITAL
rings as desirable SUBSCRIBER ACCESS
topologies. NGDLC systems that support advanced interfaces or topologies
necessarily use a COT to separate out the special services from the POTS. In essence, the COT
performs cross-connect operations to groom and distribute various types of traffic.
FIBER IN THE LOOP
Fiber in the loop (M I L) is a generic term that refers to one of three more specific de-
scriptions of the use of fiber for local distribution. The first category is fiber to the cabinet or fiber
to the neighborhood. These systems are often installed by local telephone companies as part of the
feeder portion of their local distribution facilities. Traditional copper pairs for voice service extend
from the cabinet to subscriber residences. Because a relatively long copper pair is used for “the last
mile,” they do not provide much opportunity for enhanced service offerings. Their use and justifica-
tion are strictly based on reducing the cost of POTS distribution.
The second category of FTTL is a fiber-to-the-curb (FTTC) system. As the name implies,
these systems are designed to reach within 1000 feet of a subscriber residence. An FTTC system is
generally intended to provide enhanced services such as video or high-speed data using ADSL or
VDSL. Distribution of the enhanced services from the “curb” location is carried over wire pairs or
coaxial cable. These systems are essentially identical to advanced DLC systems with optical
transport.
The third category of FITL is fiber to the home (FTTH). These systems obviously offer
opportunities for extremely large bandwidths to the home but have significant deployment
obstacles. First, installation in established neighborhoods is expensive because underground
installations (under streets and driveways) are normally required. Second, providing network power
to subscriber equipment is a major problem. The lack of power for enhanced service applications is
not much of a consideration but power to telephones is. Local exchange carriers go to great efforts
to ensure independence from commercial power systems for both their office switching systems and
the connected subscribers. (A central office typically maintains enough batteries to keep a system
up for 24 hours if commercial power is lost. If power is out for longer than this, diesel generators
are available.) A further complication of providing telephone service over FTTH is the need to
convert a digital voice channel to analog for interfacing to conventional telephones—an additional
expense and power problem.

19
HYBRID FIBER COAX SYSTEMS
The cable TV systems installed around the country can be augmented with downstream data
transmission to subscribers by merely adding “cable modems” utilizing unused or displaced TV
channel bandwidth. Upstream transmission from the subscriber to a cable
11.6 VOICEBAND TV15head end is much
MODEMS

more difficult. Although many cable TV systems were designed and installed with upstream
transmission as an option, the bandwidth available to the upstream channels is generally limited and
often subject to very high noise and interference levels. Upstream transmission limitations can be
somewhat alleviated by utilizing a conventional telephone connection with voiceband modems with
data rates up to 28.8 kbps. The telephone modem connection is used in the same basic manner as the
upstream channel of an ADSL while relative high- bandwidth downstream data are carried on the
cable. The disadvantages of this solution include the need for a subscriber telephone line, the cost of
large numbers of telephone channel connections into a service provider, and the need to coordinate
dialup telephone connections with particular cable channel users.
The basic configuration of a hybrid fiber coax (HFC) system is depicted in Figure. The coaxial
cable portions of CATV systems are configured as tree-and-branch topologies with all customers
receiving the same, multichannel broadcast signal. Amplifiers are inserted wherever the signal level
gets unacceptably low from attenuation and branching loses. Bidirectional amplifiers are shown in
Figure under the assumption that this system is an application with a return path from the residences.
In TV-only applications the return path (if there is one) is used for premium channel selection. In
expanded service applications the return path carries voice or data with frequency division
multiplexed cable modems in a band from 5 to 42 MHz.
The optical fiber transmission portion of an HFC system represents a replacement of relatively
long haul coaxial cable sections with numerous amplifiers. For this reason, the optical transmitters
and receivers are designed to carry a wideband analog signal. Notice that the optical links are shared
by a large number of customers (anywhere

Hybrid fiber coax system configuration.

15
from 100 to 1500). Upgrading an HFC system for new services typically requires greater
penetration of the optical fiber portions so that fewer households are connected to a common
coaxial cable segment. In the limiting situation, wherein each household is connected through a
16 DIGITAL
dedicated coaxial cable,SUBSCRIBER
an HFC ACCESS
system becomes an FTTC system.
Downstream digital services can utilize cable modems that typically pack 30-40 Mbps into a
6-MHz analog TV channel. 64-QAM modulation is commonly used. In newer HFC systems, new
downstream digital services can be carried at frequencies above 450 MHz while the band from 54 to
450 MHz is reserved for traditional analog TV.
A major impediment to upgrading an HFC system for return channel services is the shared
use of a coaxial cable segment common to some number of households. The network termination
within each home is passive and bidirectional, which means that all noise and interference within a
home is passed onto the common cable to all other homes. Thus, a single source of interference can
disrupt the signal to all other homes served by the common coaxial cable. Furthermore, the noise
and interference of all households are additive, indicating the need to limit the number of
households served by a single coax segment. An additional drawback of the shared cable is the need
for some form of encryption for content security. TDMA return channels help minimize the
interference problem by blocking all output energy from a residence except when an allotted time
slot occurs.
Using an HFC system for POTS has the same basic drawbacks as an FTTH system in that
there is no inherent facility for line powering the telephones. Thus, HFC might not be used for
primary (“lifeline”) POTS distribution but could be quite effective in providing secondary telephone
applications. The main attribute of HFC for enhanced services is the ability to provide dynamic
assignment of high-rate digital downstream channels and relatively low rate full-period upstream
channels.

VOICEBAND MODEMS
Voiceband modem technology improved dramatically in the early 1990s with the cul-
mination of 33.6 kbps becoming standard with ITU recommendation V.34, The rapid advance of
voiceband modem performance was due to two primary factors: the availability of economical DSP
technology for equalization/echo canceling and the improved quality of the network in terms of
lower noise and distortion resulting from the near-all-digital implementation. In an all-digital
network the only significant source of noise is the quantization noise of the analog-to-digital
conversion.

19
PCM Modems
A V.34 voiceband modem provides data rates that are near the theoretical limit imposed by
quantization noise alone. Recognition that the principal source of noise in the end-to-end connection
is the quantization noise of the A/D converters leads to alternative modem
11.6 VOICEBAND implementations
MODEMS 17 that
directly utilize the digital 64-kbps channel and eliminate the quantization noise. These modems are
commonly referred to as PCM modems.
As shown in Figure, a V.90 PCM modem relies on the digital network to carry an unaltered
digital signal from a digital source to a digital-to-analog conversion device (codec) at an analog
subscriber interface. The codec converts the PCM codewords to PAM samples that are detected by
the receiving customer premises modem and converted back to the original PCM data. Successful
data detection requires the receiving circuitry to adequately equalize the combined distortion of the
D/A smoothing filter and the transmission link, to know the quantization levels of the codec, and to
become synchronized to the D/A conversion clock. The equalization and quantization requirements
are determined during an initialization process while clock synchronization requires processing of
data transitions in the multilevel received waveform.
It might seem that a PCM modem could provide a data rate of 64 kbps. However, several
factors restrict the data rate to something less than 64 kbps. The first of these is the bandpass filter in
the D/A codec (for 60 Hz elimination and sample smoothing). A second constraint is the possibility
that the digital path through the network might

Downstream V.90 modem concept

17
include a digital pad for changing the signal level of the assumed analog signal/ A third constraint is
the possibility (in North America) that robbed bit signaling might be in use on one or more of the
18 DIGITAL SUBSCRIBER ACCESS
digital links.
The fact that the overall bandwidth of the channel is slightly over 3kHz implies that the
maximum, intersymbol, interference-free pulse rate is just over 6 kHz. Thus, the actual sample rate
of 8 kHz implies that some amount of intersymbol interference is inevitable (assuming the PCM
samples are independent).
The lack of a low-frequency transmission response is accommodated by V.90 modems by
utilizing every eighth PCM sample solely for dc restoration. Thus, the maximum data rate is 56
kbps. If robbed bit signaling is present, its effects are minimized by determining, during
initialization, which frames in the six frame sequences are signaling frames and then using only 7
bits per sample in those frames and 8 bits per sample in the nonsignaling frames. Digital pads can be
accommodated by detecting their presence and modifying the digital codewords according to the
particular amount of attenuation being inserted by the pad.
The V.90 uplink direction is implemented as a conventional (V.34) modem, thereby implying
asymmetry in the data rates. It is conceivable that PAM signaling could be used in both directions,
but the uplink is more complicated to implement and is often unnecessary because most applications
(e.g., Internet access) are inherently asymmetric in the data rates required. It is also possible to
utilize analog PAM on both ends of the connection (as opposed to just one end), but these
implementations are considerably more complicated.
The main attraction of PCM modems is that they provide almost the same data rate as an
ISDN B channel but do not require changes in the line interfaces or special treatment of the
customer loop (e.g., the elimination of bridged taps). When the customer loop is a short drop from a
remote terminal of a digital loop carrier, the maximum data rate of 53-56 kbps is assured.
LOCAL MICROWAVE DISTRIBUTION SERVICE
The FCC has allocated two separate microwave frequency bands for digital radio access
communications services. The first of these is Multichannel Multipoint Distribution Service
(MMDS), which operates at 2 GHz. MMDS is essentially a wireless cable system and, as such,
provides only one-way transmission. Reverse channel communications requires dial-up modem
connections through the telephone network.
Local Microwave Distribution Service (LMDS) operates at 28 GHz and provides cell-based,
two-way communications. Bandwidths available with LMDS are dependent on a particular service
provider’s frequency plan. Typically, these plans provide bidirectional data rates on the order of 1.5-

19
50 Mbps, although the total allocated bandwidth of almost 1 GHz can be partitioned to provide
higher rates or asymmetric rates if desired. International LMDS allocations have been made in a
range of 25-31 GHz. These systems are also referred to as Local Multipoint Communications
2 DIGITAL SUBSCRIBER ACCESS
Systems (LMCS) in Canada and Brazil.
The principal applications supported by LMDS are voice, video, and high-bandwidth data.
The immense amount of bandwidth available is the main attraction. The major cost of an LMDS
system is associated with the subscriber radio equipment. Thus, after the cell sites are established,
additional costs are incurred only as subscribers begin paying for service. This situation is in contrast
with new fiber-based facilities that require major, up-front investment before revenues are realized.
The most immediate opportunity for LMDS is to offer an alternative to high-speed digital leased
lines for businesses. High-speed Internet access and HDTV to individual residences are secondary
opportunities.
The major disadvantages of LMDS are the need for FCC licensing, line-of-sight
transmission, distance limits of 2-3 miles, extreme attenuation in heavy rainfalls, potential
interference from other services such as satellites in the same bands, and the need for local power in
all subscriber (telephone) sites. Fade margins or dynamic power control on the order of 40 dB
overcome all but the most severe rainfall rates.
The requirement for line-of-sight transmission means that a single base station cannot
communicate with all locations in a cell (except maybe in West Texas or the Australian outback).
Overlapping cells provide greater coverage, but some locations are sure to be in the shadow of all
base stations unless reflectors can be installed. Line-of- sight transmission can be further
compromised by the appearance of buildings or trees after initial deployment.
DIGITAL SATELLITE SERVICES
Direct Satellite Service (DSS) has recently become a viable alternative to cable TV services.
Because the DSS system utilizes digital transmission for its TV channels, it is straightforward for a
DSS provider to offer data communications services to the TV subscribers. However, because DSS
is a one-way communications service, dial-up modem connections through the telephone network
are required for two-way data communications. DSS is also limited to the total bandwidth (400
MHz) of the satellite transponders, which must be shared by the TV transmissions and any
allocated data channels.
In contrast to DSS, which uses geostationary satellite orbits, Low Earth Orbit Satellites of
Iridium and Teledesic provide opportunities for lower cost, bidirectional user terminals

19