MULTIPLEXING

MULTIPLEXING

Typically, two communicating stations will not utilize the full capacity of

a data link. For efficiency, it should be possible to share that capacity. A generic

term for such sharing is multiplexing.

A common application of multiplexing is in long-haul communications. Trunks

on long-haul networks are high-capacity fiber, coaxial, or microwave links. These

links can carry large numbers of voice and data transmissions simultaneously using

multiplexing.

Figure 7.1 depicts the multiplexing function in its simplest form. There are n

inputs to a multiplexer. The multiplexer is connected by a single data link to a

demultiplexer. The link is able to carry n separate channels of data. The multiplexer

combines (multiplexes) data from the n input lines and transmits over a highercapacity

data link. The demultiplexer accepts the multiplexed data stream, separates

(demultiplexes) the data according to channel, and delivers them to the appropriate

output lines.

The widespread use of multiplexing in data communications can be explained

by the following:

1. The higher the data rate, the more cost-effective the transmission facility. That

is, for a given application and over a given distance, the cost per kbps declines

with an increase in the data rate of the transmission facility. Similarly, the cost

of transmission and receiving equipment, per kbps, declines with increasing

data rate.

2. Most individual data-communicating devices require relatively modest datarate

support. For example, for most terminal and personal computer applications,

a data rate of between 9600 bps and 64 kbps is generally adequate.

The preceding statements were phrased in terms of data communicating

devices. Similar statements apply to voice communications; that is, the greater the

capacity of a transmission facility, in terms of voice channels, also, the less the cost

per individual voice channel; so, the capacity required for a single voice channel is

modest.

This lesson concentrates on three types of multiplexing techniques. The first,

frequency-division multiplexing (FDM), is the most heavily used and is familiar to

anyone who has ever turned on a radio or television set. The second is a particular

case of time-division multiplexing (TDM) known as synchronous TDM. This is

commonly used for multiplexing digitized voice streams and data streams. The third

type seeks to improve on the efficiency of synchronous TDM by adding complexity

to the multiplexer. It is known by a variety of names, including statistical TDM,

asynchronous TDM, and intelligent TDM. This lesson uses the term statistical TDM,

which highlights one of its chief properties.

7.1 FREQUENCY-DIVISION MULTIPLEXING

Characteristics

FDM is possible when the useful bandwidth of the transmission medium exceeds

the required bandwidth of signals to be transmitted. A number of signals can be carried

simultaneously if each signal is modulated onto a different carrier frequency

and the carrier frequencies are sufficiently separated that the bandwidths of the signals

do not overlap. A general case of FDM is shown in Figure 7.2a. Six signal

sources are fed into a multiplexer, which modulates each signal onto a different frequency

(f1, . . . , f6). Each modulated signal requires a certain bandwidth centered

around its carrier frequency, referred to as a channel. To prevent interference, the

channels are separated by guard bands, which are unused portions of the spectrum.

The composite signal transmitted across the medium is analog. Note, however,

that the input signals may be either digital or analog. In the case of digital input, the

input signals must be passed through modems to be converted to analog. In either

case, each input analog signal must then be modulated to move it to the appropriate

frequency band.

A familiar example of FDM is broadcast and cable television. The television

signal discussed in Lesson 2 fits comfortably into a 6-MHz bandwidth. Figure 7.3

depicts the transmitted TV signal and its bandwidth. The black-and-white video signal

is AM modulated on a carrier signal fcv. Because the baseband video signal has

a bandwidth of 4 MHz, we would expect the modulated signal to have a bandwidth

of 8 MHz centered on fcv. To conserve bandwidth, the signal is passed through a

sideband filter so that most of the lower sideband is suppressed. The resulting signal

extends from about fcv - 0.75 MHz to fcv + 4.2 MHz. A separate color subcarrier,

fv, is used to transmit color information. This is spaced far enough from f,,

that there is essentially no interference. Finally, the audio portion of the signal is

modulated on f,,, outside the effective bandwidth of the other two signals. A bandwidth

of 50 kHz is allocated for the audio signal. The composite signal fits into a

6-MHz bandwidth with the video, color, and audio signal carriers at 1.25 MHz,

4.799545 MHz, and 5.75 MHz, respectively, above the lower edge of the band. Thus,

multiple TV signals can be frequency-division multiplexed on a CATV cable, each

with a bandwidth of 6 MHz. Given the enormous bandwidth of coaxial cable (as

much as 500 MHz), dozens of TV signals can be simultaneously carried using FDM.

Of course, using radio-frequency propagation through the atmosphere is also a

form of FDM; Table 7.1 shows the frequency allocation in the United States for

broadcast television.

Let us consider a simple example of transmitting three voice signals simultaneously

over a medium. As was mentioned, the bandwidth of a voice signal is generally

taken to be 4 kHz, with an effective spectrum of 300 to 3400 Hz (Figure 7.5a).

If such a signal is used to amplitude-modulate a 64-kHz carrier, the spectrum of Fig

ure 7.5b results. The modulated signal has a bandwidth of 8 kHz, extending from 60

to 68 kHz. To make efficient use of bandwidth, we elect to transmit only the lower

sideband. Now, if three voice signals are used to modulate carriers at 64, 68, and

72 kHz, and only the lower sideband of each is taken, the spectrum of Figure 7.5~

results.

This figure points out two problems that an FDM system must cope with. The

first is crosstalk, which may occur if the spectra of adjacent component signals overlap

significantly. In the case of voice signals, with an effective bandwidth of only

3100 Hz (300 to 3400), a 4-kHz bandwidth is adequate. The spectra of signals produced

by modems for voiceband transmission also fit well in this bandwidth.

Another potential problem is intermodulation noise, which was discussed in Lesson

2. On a long link, the nonlinear effects of amplifiers on a signal in one channel

could produce frequency components in other channels.

Analog Carrier Systems

The long-distance carrier system provided in the United States and throughout the

world is designed to transmit voiceband signals over high-capacity transmission

links, such as coaxial cable and microwave systems. The earliest, and still most common,

technique for utilizing high-capacity links is FDM. In the United States,

AT&T has designated a hierarchy of FDM schemes to accommodate transmission

systems of various capacities. A similar, but unfortunately not identical, system has

been adopted internationally under the auspices of ITU-T (Table 7.2).

At the first level of the AT&T hierarchy, 12 voice channels are combined to

produce a group signal with a bandwidth of 12 X 4 kHz = 48 kHz, in the range 60

to 108 kHz. The signals are produced in a fashion similar to that described above,

using subcarrier frequencies of from 64 to 108 kHz in increments of 4 kHz. The next

basic building block is the 60-channel supergroup, which is formed by frequencydivision

multiplexing five-group signals. At this step, each group is treated as a

single signal with a 48-kHz bandwidth and is modulated by a subcarrier. The

subcarriers have frequencies from 420 to 612 kHz in increments of 48 kHz. The

resulting signal occupies 312 to 552 kHz.

There are several variations to supergroup formation. Each of the five inputs

to the supergroup multiplexer may be a group channel containing 12 multiplexed

voice signals. In addition, any signal up to 48 kHz wide whose bandwidth is contained

within 60 to 108 kHz may be used as input to the supergroup multiplexer. As

another variation, it is possible to directly combine 60 voiceband channels into a

supergroup; this may reduce multiplex costs where an interface with existing-group

multiplex is not required.

The next level of the hierarchy is the mastergroup that combines 10 supergroup

inputs. Again, any signal with a bandwidth of 240 kHz in the range 312 to

552 kHz can serve as input to the mastergroup multiplexer. The mastergroup has a

bandwidth of 2.52 MHz and can support 600 voice-frequency (VF) channels.

Higher-level multiplexing is defined above the mastergroup, as shown in Table 7.2.

Note that the original voice or data signal may be modulated many times. For

example, a data signal may be encoded using QPSK to form an analog voice signal.

This signal could then be used to modulate a 76-kHz carrier to form a component

of a group signal. This group signal could then be used to modulate a 516-kHz carrier

to form a component of a supergroup signal. Each stage can distort the original

data; this is so, for example, if the modulator/multiplexer contains nonlinearities or

if it introduces noise.