Lecture-08

burpfancyElectronics - Devices

Nov 8, 2013 (3 years and 9 months ago)

65 views

CSN200 I
ntroduction to Telecommunications,
Winter 2000


Lecture
-
08
Data Link Layer

burpfancy_e5aa8276
-
2b0f
-
4821
-
9afc
-
8c42350279d2.doc


Page
1

(
6
)

The Data Link Layer

[Ref: Chap
-
5 Fitzgerald & Dennis, Chap
-
5 p.142 Stallings]

The data link layer accepts messages from the network layer. It controls the way messages are
sent on the physical media. This layer detects and corrects error.

It also receive
s streams of bits from the physical layer and organizes them into messages that it
passes to the Network layer.

Both the sender and receiver have to agree on the rules or
protocols
that govern how they will
communicate with each other.

A
data link proto
col
determines:



who can transmit at any given time (Media Access Control)



where a message begins and ends (message framing)



how a receiver recognizes and corrects a transmission error (error control).



Media Access Control:

Media access c
ontrol refers to the method used to control
when

devices transmit data. It
becomes necessary when devices
share a common medium or circuit
, such as a multi
-
point configuration (p.135). It is very important in LANs where many users share the
same cable.

T
here is
no
need of media access control on a point
-
to
-
point full
-
duplex configuration.

Ethernet LAN uses a kind of media access control called "contention" means
competition.

Application
Network
Data Link
Physical
Application
Network
Data Link
Physical

CSN200 I
ntroduction to Telecommunications,
Winter 2000


Lecture
-
08
Data Link Layer

burpfancy_e5aa8276
-
2b0f
-
4821
-
9afc
-
8c42350279d2.doc


Page
2

(
6
)



Controlled Access


In this mode, access is usually controlled by a host mainframe c
omputer (or its front
-
end
processor).



X
-
ON/X
-
OFF

An old technique used for transmission of text messages between computers or
between a computer and a printer. X
-
OFF and X
-
ON are simply two ASCII control
characters (corresponding to ^S and ^Q on the keybo
ard) that are used to STOP and
START transmission of data. Its simplicity makes it subject to error and confusion
and is rarely used for this purpose now. Because the X
-
OFF and X
-
ON signals can be
easily lost during transmission.



POLLING ( See Fig 6
-
2)

Po
lling is the process of sending a signal to a terminal, giving it permission to transmit
or asking it to receive data. Typically a mainframe computer polls all of the connected
terminals in some order (roll call polling), giving each a turn to send or rec
eive.
Priority can be granted simply by polling high
-
priority stations more often. Polling
involves some delay in servicing individual terminals.




Contention


Devices
contend

or compete for the use of the medium. (This is used in Ethernet LANs,
for exam
ple, where all PCs share the use of the same cable.)

Devices wait until the circuit is free (no one else is using it) and then transmit whenever
they have data to send. This technique works well under low to medium usage but too
many
collisions
occur if u
sage is high.




Polling


Polling is the process of sending a permission signal to transmit or to receive. This is
analogous to classroom situation where the teacher decides who will answer the question
out of many students.

There are several types of pollin
g. With roll call polling each client computer is polled
sequentially. Priority can be given to some clients by polling more often than others. Hub
polling often called token passing is used on a LAN that does not have a central
computer. Here one computer

starts the poll and passes the poll to the next computer.
Token ring LAN is an example of hub polling.

Which one is better?

In general, contention approaches work better than controlled for small networks. In high
volume networks, a well controlled protoc
ol prevents collisions.

CSN200 I
ntroduction to Telecommunications,
Winter 2000


Lecture
-
08
Data Link Layer

burpfancy_e5aa8276
-
2b0f
-
4821
-
9afc
-
8c42350279d2.doc


Page
3

(
6
)

Data Link Protocols:


These protocols transmit error
-
free data from one computer to another.

The data is grouped into blocks or packets so transmission is more efficient and error
-
checking
can be performe
d easily.

All of the ones shown below use
Controlled Media Access
methods.


Protocol

Size of
Blocks
(bytes)

Error Detection

Retransmission

Media Access &
Characteristics

Asynchronous
Transmission

1

1
-
bit parity

Continuous ARQ

Full Duplex;

Very inefficien
t

File Transfer
Protocols





XMODEM

132

8
-
bit checksum

Stop
-
and
-
Wait ARQ

Controlled Access

XMODEM
-
CRC

132

8
-
bit CRC

Stop
-
and
-
Wait ARQ

Improved error checking

XMODEM
-
1K

1028

8
-
bit CRC

Stop
-
and
-
Wait ARQ

More efficient because
of larger block size.

YMOD
EM

1029

16
-
bit CRC

Stop
-
and
-
Wait ARQ

Improved error checking

ZMODEM

varies

32
-
bit CRC

Continuous ARQ

An excellent file transfer
protocol

KERMIT

varies

24
-
bit CRC

Continuous ARQ

Kermit has been greatly
improved.

Typically uses 1000
-
byte packets.

Synchron
ous
Protocols





SDLC

varies

16
-
bit checksum

Continuous ARQ

Controlled Access

HDLC

varies

16
-
bit CRC

Continuous ARQ

Controlled Access

Token ring

varies

32
-
bit CRC

Stop
-
and
-
Wait ARQ

Controlled Access

Ethernet

varies

32
-
bit CRC

Stop
-
and
-
Wait ARQ

Content
ion

SLIP

varies

None

None

Full Duplex

PPP

varies

16
-
bit CRC

Continuous ARQ

Full Duplex



Asynchronous Transmission:


Also known as start
-
stop transmission. It is typically used on point
-
to
-
point full
-
duplex circuits,
so
media access control

is not a con
cern. The start and stop bits (
message delineation
) does
the synchronization (when to sample the bit at the receiver end) between the transmitter and
the receiver.
Error detection

is achieved by using a parity bit for each character.

It is simple and chea
p but requires an overhead more than 22% (=100x7/9). Thus transmission
efficiency is low, 78% or less.


CSN200 I
ntroduction to Telecommunications,
Winter 2000


Lecture
-
08
Data Link Layer

burpfancy_e5aa8276
-
2b0f
-
4821
-
9afc
-
8c42350279d2.doc


Page
4

(
6
)

Asynchronous File Transfer Protocols:


Also used on point
-
to
-
point full
-
duplex circuits. Can also be used on a half
-
duplex circuit using
controlled acc
ess.

To increase the efficiency, blocks of data are grouped together and then transmitted, rather than
sending each character individually.

Instead of individual start and stop bits for each character these protocols use a start of header
character (
messa
ge delineation
), 1 byte block number, 128/1024 character block, and a 1
-
byte checksum/CRC (cyclic redundancy check) for
error checking
. (Fig 5
-
9). For error
correction these protocols use ARQ (automatic repeat request).

XMODEM, YMODEM, ZMODEM and KERMIT ar
e the examples.



Synchronous Transmission:


It is used on
shared circuits

and uses some kind of media access protocols.


In Synchronous Transmission, a large group of data is transmitted together as a
frame
or
packet
just as the Asynchronous File Transfer

Protocols. To prevent
timing drift

between transmitter
and receiver, their clocks must somehow be synchronized. This timing drift is not a serious
problem in asynchronous file transfer protocols, as they use lower transmission rate (0.56
Mbps) compared to

the synchronous transmission (10Mbps or higher).

One possibility is to provide a separate clock line between transmitter and receiver. This
technique works well for a short distance but over a long distance clock pulse itself are
subject to transmission i
mpairments and timing errors can occur.

The other alternative is to embed the clocking information in the data signal; for digital
transmission, this can be accomplished with Manchester or Differential Manchester encoding.
For analog signals, the carrier
frequency itself can be used to synchronize the receiver.


The
beginning and end of each frame

are
marked
in some special way so the receiver will be
able to identify when a frame begins and ends, another kind of synchronization. This is
different from
as
ynchronous

transmission where each character being sent had a start and stop
character.


They also use some kind of error control techniques, mostly CRC.



Synchronous Data Link Protocols:



determine who can transmit at any given time (Media Access Control)



decide where frames begin and end and how fields within the frame will be interpreted
(message framing)



detect transmission errors.


CSN200 I
ntroduction to Telecommunications,
Winter 2000


Lecture
-
08
Data Link Layer

burpfancy_e5aa8276
-
2b0f
-
4821
-
9afc
-
8c42350279d2.doc


Page
5

(
6
)

Three Categories of Synchronous Data Link Protocols:



Byte
-
oriented protocols
-

IBM’s
Token Ring



Bit
-
oriented protocols

-

IBM’s
SDLC

and ISO’s

HDLC



Byte
-
count protocols
-

DEC’s
DECNET,

Ethernet
(IEEE 802.3)



Synchronous Data Link Control (SDLC):
(Fig 5
-
10)


IBM bit
-
oriented protocol.

The data or message can be any number of
bits

A unique
flag
(01111110) is used at t
he beginning and end of all frames.

It uses a special
bit stuffing
technique to insert 0’s after any consecutive 5 1’s in a row in the
message so the uniqueness of the special flag is not lost. The receiver automatically deletes a
0 after any 5 consecutiv
e 1’s. This solves the
transparency problem.

Uses
CRC
-
16

for
error detection

and
Continuous
-
ARQ

for
error correction
.

Used in IBM's proprietary networks to connect remote terminals to mainframes.

The sender and receiver are required to be
synchronized

bef
ore transmission.

It uses a controlled
-
access media access protocol.



High
-
Level Data Link Control (HDLC):


ISO’s bit
-
oriented protocol.

Essentially the same as SDLC.

It uses a controlled
-
access media access protocol.



Token Ring:


Token Ring is used in
Local Area Networks, developed by IBM, IEEE standard is 802.5.

It is a byte
-
oriented protocol that does not suffer the transparency problem as SDLC. The start
and end delimiters are special electrical signal different from any other bit pattern and thus
ca
n not be confused with data bits.

Message field is variable and maximum is 4500 bytes.

Uses 32
-
bit CRC.



Ethernet:


Ethernet

is used in Local Area Networks, developed by Digital, Intel and Xerox, IEEE standard
is 802.3.

It is a byte
-
count protocol. Instea
d of using special characters or bit patterns to mark the end of
packet, it includes a field that specifies the length of the message portion of the packet, thus
has no transparency problem.

CSN200 I
ntroduction to Telecommunications,
Winter 2000


Lecture
-
08
Data Link Layer

burpfancy_e5aa8276
-
2b0f
-
4821
-
9afc
-
8c42350279d2.doc


Page
6

(
6
)

Message field is variable and maximum is 1500 bytes.

The packet e
nds with a 32
-
bit CRC code.


It is a byte
-
oriented protocol that does not suffer the transparency problem as SDLC. The start
and end delimiters are special electrical signal different from any other bit pattern and thus
can not be confused with data bits.

Message field is variable and maximum is 4500 bytes.

Uses 32
-
bit CRC.



SLIP
(Serial Line Internet Protocol)
:


It is a byte
-
oriented protocol that suffers the transparency problem because of the end characters
(11000000).

No addressing.

No error control.



PPP
(Point
-
to
-
Point Protocol)
:


It is a byte
-
oriented protocol that suffers the transparency problem because of the end characters
(01111110).

It is replacement of SLIP.

CRC
-
16 error control.

No media access because it is a point
-
to
-
point protocol.



Tran
smission Efficiency:


Transmission Efficiency is defined as the total number of information bits divided by the total
bits in transmission (i.e., information bits plus the overhead bits).


Synchronous networks (>90%) usually are more efficient than asynchr
onous networks (<78%).

For an Ethernet, the transmission efficiency is 98.3% (=1500 bytes/1526 bytes).


In general the larger the message field, the higher the efficiency.

So, why not have 10k or even 100k packets to really increase the efficiency?

The ans
wer is that anytime a packet is received containing an error, the entire packet must be
retransmitted.

The probability that a packet contains an error increases with the size of the packet. (4k~8k)



Throughput (TRIB, Transmission Rate of Information Bits)
:


Throughput is the total number of information bits received per second, after taking account the
overhead bits and the need to retransmit packets containing errors.

TRIB is the
effective rate of data transfer

(TRIB = # of bits accepted/Time).