COPPERSTONE UNIVERSITY

BCS/BIT: DATA COMMUNICATIONS AND NETWORKING

Long Distance Communication (Carriers, Modulation and Modems)

Introduction

Electric current cannot be propagated an arbitrary distance over copper wire because the current
becomes weaker as it travels. The resistance in the wire causes small amounts of the electrical
energy to be converted to heat. Current is therefore lost as heat (there are other losses as well).
The resulting signal loss may prevent accurate decoding of data. Signal loss prevents use of RS-
232 over long distances. In long-distance communication, a continuous, oscillating signal is
used. A continuous, oscillating signal will propagate further than other signals. This continuous
oscillating signal, usually a sine wave is called a carrier. Figure 1 illustrates a carrier waveform.

Figure 1: The waveform of a typical carrier. The carrier oscillates continuously, even when no data is sent.

To send data, the transmitter modifies the carrier; this is called modulation and was originally
devised for use with radio and television. This technique is used over a variety of media, e.g.
wire, optical fibres, microwaves, radio frequencies, etc. The transmitter generates a carrier signal
which it modulates with the data being sent. The receiver monitors the incoming carrier, detects
modulation, and reconstructs the original data.

Multiplexing and Demultiplexing (Channelization)

The Concept of Multiplexing

We use the term multiplexing to refer to the combination of information streams from multiple
sources for transmission over a shared medium, and multiplexor to denote a mechanism that
implements the combination. Similarly, we use the term demultiplexing to refer to the separation
of a combination back into separate information streams, and demultiplexor to refer to a
mechanism that implements separation. Multiplexing and demultiplexing are not restricted to
hardware or to individual bit streams — the idea of combining and separating communication

1
forms a fundamental basis used in many parts of computer networking. Figure 2 illustrates the
concept.

Figure 2 the concept of multiplexing in which independent pairs of senders and receivers share a transmission medium.

In the figure, each sender communicates with a single receiver. Although they carry on
independent communication, all pairs share a single transmission medium. The multiplexor
combines information from the senders for transmission in such a way that the demultiplexor can
separate the information for receivers.

THE BASIC TYPES OF MULTIPLEXING

There are four basic approaches to multiplexing that each has a set of variations and
implementations.

 Frequency Division Multiplexing

 Wavelength Division Multiplexing
 Time Division Multiplexing
 Code Division Multiplexing

Time and frequency division multiplexing are widely used. Wavelength division
multiplexing is a form of frequency division multiplexing used for optical fiber. Code division
multiplexing is a mathematical approach used in cell phone mechanisms.

FREQUENCY DIVISION MULTIPLEXING (FDM)

Frequency Division Multiplexing (FDM) is easy to understand because it forms the basis for
broadcast radio. The underlying principle arises from the physics of transmission: a set of radio
stations can transmit electromagnetic signals simultaneously without interference provided they
each use a separate channel (i.e., carrier frequency). Data communications systems apply the
same principle by simultaneously sending multiple carrier waves over a single copper wire or
using wavelength division multiplexing to send multiple frequencies of light over an optical
fiber. At the receiving end, a demultiplexor applies a set of filters that each extracts a small range
of frequencies near one of the carrier frequencies. Figure 3 illustrates the organization.

2
Figure 3 Illustration of the basic FDM demultiplexing where a set of filters each selects the frequencies for one channel
and suppresses other frequencies.

A key idea is that the filters used in FDM only examine frequencies. If a sender and receiver
pair are assigned a particular carrier frequency, the FDM mechanism will separate the frequency
from others without otherwise modifying the signal. Thus, any of the modulation techniques
discussed in the earlier can be used with any carrier.
The most significant advantage of FDM arises from the simultaneous use of a transmission
medium by multiple pairs of communicating entities. We imagine FDM as providing each pair
with a private transmission path as if the pair had a separate physical transmission medium.

Figure 4 The conceptual view of FDM as providing a set of independent channels

Any practical FDM system imposes limits on the set of frequencies that can be used for
channels. If the frequencies of two channels are arbitrarily close, interference can occur.
Furthermore, demultiplexing hardware that receives a combined signal must be able to divide the
signal into separate carriers. Data communications systems, designers ensure that there is
adequate spacing between carrier frequencies. The gap between carrier frequencies is known as a
guard band.

As an example of channel allocation, consider the assignment in Figure 5 that allocates 200
KHz to each of 6 channels with a guard band of 20 KHz between each.

3
 Figure 5 Frequency allocation illustration in FDM

USING A RANGE OF FREQUENCIES PER CHANNEL

Most FDM systems assign each sender and receiver pair a range of frequencies and the
ability to choose how the frequencies can be used. There are two primary ways that systems use
a range of frequencies.

 Increase the data rate

 Increase immunity to interference

To increase the overall data rate, a sender divides the frequency range of the channel into K
carriers, and sends 1/K of the data over each carrier. In essence, a sender performs frequency
division multiplexing within the channel that has been allocated. Some systems use the term
subchannel allocation to refer to the subdivision.

To increase immunity to interference, a sender uses a technique known as spread

spectrum1. Various forms of spread spectrum can be used, but the basic idea is to divide the
range of the channel into K carriers, transmit the same data over multiple channels, and allow a
receiver to use a copy of the data that arrives with fewest errors. The scheme works extremely
well in cases where noise is likely to interfere with some frequencies at a given time.

Hierarchical FDM

Some of the flexibility in FDM arises from the ability of hardware to shift frequencies. If a
set of incoming signals all use the frequency range between 0 and 4 KHz, multiplexing hardware
can leave the first stage as is, map the second onto the range 4 KHz to 8 KHz, map the third onto
the range 8 KHz to 12 KHz, and so on. The technique forms the basis for a hierarchy of FDM
multiplexors that each map their inputs to a larger, continuous band of frequencies. Figure 6
illustrates the concept of hierarchical FDM.

1
You are encouraged to read more on spread spectrum

4
 Figure 6 Hierarchical FDM

As the figure illustrates, the basic input consists of a set of twelve analog telephone signals,
which each occupy frequencies 0 through 4 KHz. At the first stage, the signals are multiplexed
into a single signal known as a group that uses the frequency range of 0 through 48 KHz. At the
next stage, five groups are multiplexed into a single super- group that uses frequencies 0 through
240 KHz, and so on. At the final stage, 3600 telephone signals have been multiplexed into a
single signal.

Wavelength Division Multiplexing (WDM)

The term Wavelength Division Multiplexing (WDM) refers to the application of frequency
division multiplexing to optical fiber. The inputs and outputs of such multiplexing are
wavelengths of light, denoted by the Greek letter λ, and informally called colors. To understand
how multiplexing and demultiplexing can work with light, recall from basic physics that when
white light passes through a prism, colors of the spectrum are spread out. A prism operates in the
reverse mode as well: if a set of colored light beams are each directed into a prism at the correct
angle, the prism will combine the beams to form a single beam of white light. Finally, recall that
what humans perceive as a color is in fact a range of wavelengths of light. Prisms form the basis
of optical multiplexing and demultiplexing. A multiplexor accepts beams of light of various
wavelengths and uses a prism to combine them into a single beam; a demultiplexor uses a prism
to separate the wavelengths. Figure 7 illustrates the concept.

Figure 7 Wavelength Division Multiplexing

5
Time Division Multiplexing (TDM)

The chief alternative to FDM is known as Time Division Multiplexing (TDM). TDM is less
esoteric than FDM and does not rely on special properties of electromagnetic energy. Instead,
multiplexing in time simply means transmitting an item from one source, then transmitting an
item from another source, and so on. Figure 8 illustrates the concept.

Figure 8 Time Division Multiplexing

Synchronous TDM

Time division multiplexing is a broad concept that appears in many forms and is widely
used throughout the Internet. Thus, the diagram in Figure 8 is merely a conceptual view, and the
details may vary. For example, the figure shows items being sent in a round-robin fashion (i.e.,
an item from sender 1 followed by an item from sender 2, etc). Although some TDM systems use
round-robin order, other do not. A second detail in Figure 8 does not apply to all types of TDM.
Namely, the figure shows a slight gap between items. When TDM is applied to synchronous
networks, no gap occurs between items. The result is known as Synchronous Time Division
Multiplexing. A synchronous TDM system uses round- robin order to select items. Figure 9
illustrates how synchronous TDM works for a system of four senders.

Figure 9 illustration of Synchronous TDM system with four senders

6
Framing Used In the Telephone System Version of TDM

Telephone systems use synchronous TDM to multiplex digital streams from multiple phone
calls over a single medium. In fact, telephone companies use the acronym TDM to refer to the
specific form of TDM used to multiplex digital telephone calls. The phone system standards for
TDM include an interesting technique to insure that a demultiplexor stays synchronized with the
multiplexor. To understand why synchronization is needed, observe that a synchronous TDM
system sends one slot after another without any indication of the output to which a given slot
occurs. Because a demultiplexor cannot tell where a slot begins, a slight difference in the clocks
used to time bits can cause a demultiplexor to misinterpret the bit stream. To prevent
misinterpretation, the version of TDM used in the phone system includes an extra framing
channel as input. Instead of taking a complete slot, framing inserts a single bit in the stream on
each round. Along with other channels, a demultiplexor extracts data from the framing channel
and checks for alternating 0 and 1 bits. The idea is that if an error causes a demultiplexor to lose
a bit, it is highly likely that the framing check will detect the error and allow the transmission to
be restarted. Figure 10 illustrates the use of framing bits.

Figure 10 Illustration of the synchronous TDM system used by the telephone system in which a framing bit precedes each round
of slots.

HIERARCHICAL TDM

Like FDM, TDM can be arranged in a hierarchy. The difference is that each successive stage
of a TDM hierarchy uses N times the bit rate, whereas each successive stage of an FDM
hierarchy uses N times the frequencies. Additional framing bits are added to the data, which
means that the bit rate of each successive layer of hierarchy is slightly greater than the aggregate
voice traffic. Compare the example TDM hierarchy in Figure 11 with the FDM example in
Figure 5.

7
 Figure 11 Hierarchical TDM

The Problem with Synchronous TDM: Unfilled Slots

Synchronous TDM works well if each source produces data at a uniform, fixed rate equal to
1/N of the capacity of the shared medium. For example, if a source corresponds to a digital
telephone call, the data will arrive at a uniform rate of 64 Kbps. However, many sources generate
data in bursts, with idle time between bursts, which does not work well with a synchronous TDM
system. To understand why, consider the example in Figure 12. In the figure, sources on the left
produce data items at random. Thus, the synchronous multiplexor leaves a slot unfilled if the
corresponding source has not produced an item by the time the slot must be sent. In practice, of
course, a slot cannot be empty because the underlying system must continue to transmit data.
Thus, the slot is as- signed a value (such as zero), and an extra bit is set to indicate that the value
is invalid.

Figure 12 Illustration of a synchronous TDM system leaving slots unfilled when a source does not have a data item ready
in time.

STATISTICAL TDM

How can a multiplexing system make better use of a shared medium? One technique to increase
the overall data rate is known as statistical time division multiplexing or statistical multiplexing.

8
The terminology is awkward, but the technique is straight- forward: select items for transmission
in a round-robin fashion, but instead of leaving a slot unfilled, skip any source that does not have
data ready. By eliminating unused slots, statistical TDM takes less time to send the same amount
of data. For example, Figure 13 illustrates how a statistical TDM system sends the data from
Figure 10 in only 8 slots instead of 13.

Figure 13 Statistical TDM

Although it avoids unfilled slots, statistical multiplexing incurs extra overhead. To see why,
consider demultiplexing. In a synchronous TDM system a demultiplexor knows that every N th
slot corresponds to a given receiver. In a statistical multiplexing system, the data in a given slot
can correspond to any receiver. Thus, in addition to data, each slot must contain the identification
of the receiver to which the data is being sent.

INVERSE MULTIPLEXING

An interesting twist on multiplexing arises in cases where the only connection between two
points consists of multiple transmission media, but no single medium has a bit rate that is
sufficient. At the core of the Internet, for example, service providers need higher bit rates than
are available. To solve the problem, multiplexing is used in reverse: spread a high-speed digital
input over multiple lower-speed circuits for transmission and combine the results at the receiving
end. Figure 13 illustrates the concept. In practice, an inverse multiplexor cannot be constructed
merely by connecting the pieces of a conventional multiplexor backward. Instead, hardware must
be designed so that the sender and receiver agree on how data arriving from the input will be
distributed over the lower-speed connections. More important, to insure that all data is delivered
in the same order as it arrived, the system must be engineered to handle cases where one or more
of the lower-speed connections has longer latency than others. Despite its complexity, inverse
multiplexing is widely used in the Internet.

9
 Figure 14 Inverse Multiplexing

CODE DIVISION MULTIPLEXING

A final form of multiplexing used in parts of the cellular telephone system and for some satellite
communication is known as Code Division Multiplexing (CDM). The specific version of CDM
used in cell phones is known as Code Division Multi-Access (CDMA). Unlike FDM and TDM,
CDM does not rely on physical properties, such as frequency or time. Instead, CDM relies on an
interesting mathematical idea: values from orthogonal vector spaces can be combined and
separated without interference. The particular form used in the telephone network is easiest to
understand. Each sender is assigned a unique binary code C i that is known as a chip sequence.
Chip sequences are selected to be orthogonal vectors (i.e., the dot product of any two chip
sequences is zero). At any point in time, each sender has a value to transmit, V i. The senders
each multiply Ci×Vi, and transmit the results. In essence, the senders transmit at the same time,
and the values are added together. To extract value Vi, a receiver multiplies the sum by Ci.

Modem hardware used for modulation and demodulation

A hardware mechanism that accepts a sequence of data bits and applies modulation to a carrier
wave according to the bits is called a modulator; a hardware mechanism that accepts a modulated
carrier wave and recreates the sequence of data bits that was used to modulate the carrier is
called a demodulator. Thus, transmission of data requires a modulator at one end of the
transmission medium and a demodulator at the other. In practice, most communication systems
are full duplex communication, which means each location needs both a modulator, which is
used to send data, and a demodulator, which is used to receive data. To keep cost low and make
the pair of devices easy to install and operate, manufacturers combine modulation and
demodulation mechanisms into a single device called a modem (modulator and demodulator).
Figure 15 illustrates how a pair of modems use a 4-wire connection to communicate.

10
 Figure 15 Illustration of two modems that use a 4-wire connection.

Optical and Radio Frequency Modems

Modems are also used with other media including Radio Frequency (RF) transmission and
optical fibers. A pair of RF modems can be used to send data via radio while a pair of optical
modems can be used to send data across a pair of optical fibers. Modems can use entirely
different media, but the principle remains the same:
 at the sending end, a modem modulates a carrier
 at the receiving end, data is extracted from the modulated carrier

11