|
|
Line coding standards
|
|
|
|
The "clock" simply provides
a reference for the individual binary 1s and 0s. The
idea is to have a code with regular and frequent level
transitions on the channel. The transitions delineate
the binary data cells (1s and 0s) at the receiver, and
sampling logic continuously looks for the state transitions
in order to delineate the bit streams. Receiver sampling
usually occurs at a higher rate than the data rate in
order to define the bit cells more precisely.
Below illustration provides several common binary coding
schemes used in the industry. We will discuss each of
these briefly and describe their advantages and disadvantages.
(Be aware that the signals are not as sharp and square
as the illustration below suggests.It depicts actual
signals.) All these signals exhibit one or several of
the following four Characterstics:
|
- Unipolar code. No signal
below zero voltage or no signal above (i.e., algebraic
sign does not change: 0 volts for 1 and 3 volts for
0)
- Polar code. Signal is
above and below zero voltage (opposite algebraic signs
identify logic states: +3 volts and -3 volts).
- Bipolar code. The signal
varies among three levels.
- Alternate mark inversion (AMI) code.
Uses alternate polarity pulses to encode binary 1s.
|
|
|
Digital Codes |
Above illustration shows
the non-return-to-zero code (NRZ). Notice the signal
level remains stable throughout the bit cell. In this
case, the signal level remains low for a bit 1 and goes
to a high voltage for a bit 0. (Opposite voltages are
also used in many devices.) NRZ is a widely used data
communications coding scheme because of its relative
simplicity and low cost. The NRZ code also makes very
efficient use of bandwidth, since it can represent a
bit with each baud (signal change). However, it suffers
from the lack of self-clocking capabilities, since a
long series of continuous 1s or 0s would not create
a signal state transition on the channel. As a consequence,
the receiver's clock could possibly drift from the incoming
signal and not sample the line at the right time; the
transmitter and the receiver might actually lose synchronization
with each other. The NRZ code can be polar to bipolar,
depending on the actual implementation. NRZ is widely
used in communications because it requires no encoding
or decoding and it uses a channel's bandwidth very effectively.
The return-to-zero code (RZ) usually entails the changing
of the signal state at least once in every bit cell.
This scheme is shown in graph(b) of above illustration.
Since RZ codes provide a transition in every bit cell,
they have very good synchronization characterstics.
The RZ code's primary disadvantage is that it requires
two signal transitions for each bit. Consequently, an
RZ code would require twice the baud of a conventional
NRZ code. We find this type of code in some of the more
sophisticated systems dealing with local area networks,
light wave technologies, and optic fibers.
In the above illustration, another very popular code
found in many communication systems,for example Manchester
code. This code provides a signal state in every bit
cell. Consequently, it is a good clocking code. However,
like the RZ code, it requires twice the baud for the
bit rate. In addition, the interface devices used to
achieve this higher baud are considerably more expensive
than the NRZ interfaces. Manchester code is commonly
found in magnetic tape recording, optic fiber links,
coaxial lines, and local area networks.
The above illustration also shows one code used by AT
and T, the Bell Operating Companies, and other carriers.
This was originally called the Bell System PCM Code.
This signaling structure is an example of bipolar AMI
wherein alternate polarity pulses are used to encode
logic 1. This particular code presents some problems
when a long series of zeros are located in the transmission.
The components in the system have no way to synchronize
with zero bit cells because there are no changes in
the state of the line.Now we can see there is a signal
state changes on the line periodically, even though
a long series of zeros may exist in the data stream.
Finally, above illustration shows an actual direct current
(dc) signal as it exists in the channel.
A more sophisticated process is synchronous transmission,
which uses separate clocking channels or a self-clocking
code. Synchronous formats eliminate the intermittent
start stop signals around each character. The preliminary
signals are usually called synchronization or sync bytes.
In a more modern system we call them flags or preambles.
Their principal function is to alert the receiver that
user data are arriving. This process is called framing.
It can be seen that a long synchronous data message
without intermittent start stop bits could present problems
because the receiver could drift from the signal. We
discussed in this section, two methods to deal with
this problem:
|
- provide a separate clocking channel, or
- provide a signal code that it self-clocking, such
as return-to-zero or Manchester code. The latter approach
allows the receiver to develop its timing from the
line transitions.
|
We now have enough information
on codes and formats to move to other components in
the network. However, we will come back to this subject
several times. It will be more evident in latter sections
that the knowledge of codes and signaling states is
important to gain an understanding of other aspects
of networks and protocols. |
|
|
Formats for Asynchronous,
Synchronous and Typical Synchronous Transmission. |
|
Data encoding/decoding: |
|
- Provides the signals to synchronize the stations
on the channel (this sync signal is called the preamble
);
- Encodes the binary data stream to a self clocking
code (Manchester code) at the transmitting site and
decodes the Manchester code back to binary code at
the receiver.
|
The transmit data encapsulation entity
receives the user data and constructs the MAC frame.
It also appends the frame check-sequence field to the
data and passes the frame to media-access management,
which buffers the frame until the channel is free.The
channel is sensed as free when it sees a carrier sense
signal turned off from the transmit channel-access entity
in the physical layer, after a brief delay, media-access
management passes the frame to the physical layer.
At the physical layer on the transmit side, data encoding
transmits the synchronization signal (preamble). In
addition, it encodes the binary data stream to a self-clocking
Manchester code. The signal is then passed to transmit
channel access, which introduces the signal onto the
channel.
FDDI stipulates a very unique approach to timing and
clocking on the network. The reader may recall from
previous discussions that the best code to be used in
a network is one which provides frequent signal state
changes. The changes provide the receiver with the ability
to continue to adjust to the incoming signal, thereby
assuring that the transmitting device and the receiving
device are synchronized with each other. The Manchester
code used in the IEEE 802.3 standard is only 50 percent
efficient, because every bit requires a two-state transition
on the line (i.e., two baud). Using Manchester code,
a 100-megabit transmission rate requires 200 megabits
of bandwidth (a 200 MHz rate). In other words, Manchester
code requires twice the band for its transmission. |
|
|
| Back to the
top |
|
|
|
|
|
