1. Data Link Layer
RAHUL DEVA

2. The Data Link Layer Framing Error Control Flow Control
Error Correction
Error Detection
Flow Control
Retransmission with Sliding Window Protocols
Protocol Specification and Verification
Example Data Link Protocols

3. Data Link Layer Services
Unacknowledged connectionless service
Voice
Acknowledged connectionless service
wireless
Acknowledged connection-oriented service
no out-of-order

4. Framing
To break the bit stream up into discrete frames and compute the checksum for each frame. When a frame arrives at the destination, the checksum is recomputed and validated.
One way to achieve this framing is to insert time gaps between frames. However, networks rarely make any guarantees about timing.

5. Error Control Three types of frames at a receiver’s end
Damaged frame: Negative acknowledgment / retransmission
Lost frame: Time-out mechanism
Valid frame: Sequence numbering / discarding for valid but duplicate frames

6. Flow Control
A technique for assuring that a transmitting station does not overwhelm a receiving station with data

7. Data Framing Character count
Starting and ending characters with character stuffing
Starting and ending flags with bit stuffing
Physical layer coding violations

8. Character Count
character count
frame 1
5 characters
frame 2
frame 3
8 characters
frame 4
frame 2
(wrong)
frame 1
5 characters
one-bit error
Now a character count
Even if the checksum is incorrect so the destination knows that the frame is bad, it still has no way of telling where the next frame starts.

9. Starting and Ending Characters with Character Stuffing
DLE STX user data DLE ETX
beginning
of frame
end of
frame
DLE(0x10): Data Link Escape
STX(0x02): Start of TeXt
ETX(0x03): End of TeXt
DLE STX A DLE B DLE ETX
user data 0x10
DLE stuffed at the sender
DLE destuffed at the receiver
DLE STX A DLE DLE B DLE ETX

10. Starting and Ending Flags with Bit Stuffing
user data
beginning
of frame
end of
frame
flag
bits stuffed at the sender
bits destuffed at the receiver
original data

11. Introduction: Error Detection and Correction
It is physically impossible for any data recording or transmission medium to be 100% perfect over its entire expected useful life.
Noise is always present
If a communications line experiences too much noise, the signal will be lost or corrupted
Communication systems should check for transmission errors
Once an error is detected, a system may perform some action
Some systems perform no error control, but simply let the data in error be discarded

12. Noise Also known as thermal or Gaussian noise
Relatively constant and can be reduced
If noise gets too strong, it can completely disrupt the signal

13. Noise (continued)

14. The effect of impulse noise on a digital signal
One of the most disruptive forms of noise
Random spikes of power that can destroy one or more bits of information
Difficult to remove from an analog signal because it may be hard to distinguish from the original signal
Impulse noise can damage more bits if the bits are closer together (transmitted at a faster rate)
The effect of impulse noise on a digital signal

15. Three Telephone Circuits experiencing crosstalk
Unwanted coupling between two different signal paths
For example, hearing another conversation while talking on the telephone
Relatively constant and can be reduced with proper measures
Three Telephone Circuits experiencing crosstalk

16. A signal bouncing back at the end of a cable and causing echo
The reflective feedback of a transmitted signal as the signal moves through a medium
Most often occurs on coaxial cable
If echo bad enough, it could interfere with original signal
Relatively constant, and can be significantly reduced
A signal bouncing back at the end of a cable and causing echo

17. Original digital signal and digital signal with jitter
The result of small timing irregularities during the transmission of digital signals
Occurs when a digital signal is repeated over and over
If serious enough, jitter forces systems to slow down their transmission
Steps can be taken to reduce jitter
Original digital signal and digital signal with jitter

18. Delay Distortion
Occurs because the velocity of propagation of a signal through a medium varies with the frequency of the signal
Can be reduced

19. Attenuation
The continuous loss of a signal’s strength as it travels through a medium

20. Error Prevention (continued)

21. Error Detection
Despite the best prevention techniques, errors may still happen
To detect an error, something extra has to be added to the data/signal
This extra is an error detection code
Three basic techniques for detecting errors: parity checking, arithmetic checksum, and cyclic redundancy checksum

22. Parity Checks Simple parity
If performing even parity, add a parity bit such that an even number of 1s are maintained
If performing odd parity, add a parity bit such that an odd number of 1s are maintained
For example, send using even parity
For example, send using even parity

23. Parity Checks (continued)
Simple parity (continued)
What happens if the character is sent and the first two 0s accidentally become two 1s?
Thus, the following character is received:
Will there be a parity error?
Problem: Simple parity only detects odd numbers of bits in error

24. Parity Checks (continued)
Longitudinal parity
Adds a parity bit to each character then adds a row of parity bits after a block of characters
The row of parity bits is actually a parity bit for each “column” of characters
The row of parity bits plus the column parity bits add a great amount of redundancy to a block of characters

25. Vertical Redundancy Check (VRC)
Append a single bit at the end of data block such that the number of ones is even  Even Parity (odd parity is similar)  
VRC is also known as Parity Check
Performance:
Detects all odd-number errors in a data block

26. Longitudinal Redundancy Check (LRC)
Organize data into a table and create a parity for each column
Original Data
LRC

27. Longitudinal Redundancy Check and Vertical Redundancy Check27
VRC
LRC

28. Parity Checks (continued)
Both simple parity and longitudinal parity do not catch all errors
Simple parity only catches odd numbers of bit errors
Longitudinal parity is better at catching errors but requires too many check bits added to a block of data
We need a better error detection method
What about arithmetic checksum?

29. Modulo 2
Modulo 2 arithmetic works like clock arithmetic.
In clock arithmetic, if we add 2 hours to 11:00, we get 1:00.
In modulo 2 arithmetic if we add 1 to 1, we get 0. The addition rules couldn’t be simpler:
0 + 0 = = 1
1 + 0 = = 0

30. Find the quotient and remainder when 1111101 is divided by 1101 in modulo 2 arithmetic.
As with traditional division, we note that the dividend is divisible once by the divisor.
We place the divisor under the dividend and perform modulo 2 subtraction.

31. Error Detection and Correction
Find the quotient and remainder when is divided by 1101 in modulo 2 arithmetic…
Now we bring down the next bit of the dividend.
We see that is not divisible by So we place a zero in the quotient.

32. Error Detection and Correction
Find the quotient and remainder when is divided by 1101 in modulo 2 arithmetic…
1010 is divisible by in modulo 2.
We perform the modulo 2 subtraction.

33. Error Detection and Correction
Find the quotient and remainder when is divided by 1101 in modulo 2 arithmetic…
We find the quotient is 1011, and the remainder is 0010.
This procedure is very useful to us in calculating CRC syndromes.
Note: The divisor in this example corresponds to a modulo 2 polynomial: X 3 + X

34. Cyclic redundancy checking (CRC)
Suppose we want to transmit the information string:
The receiver and sender decide to use the (arbitrary) polynomial pattern, 1101.
The information string is shifted left by one position less than the number of positions in the divisor.
The remainder is found through modulo 2 division (at right) and added to the information string: =

35. Cyclic redundancy checking (CRC)
If no bits are lost or corrupted, dividing the received information string by the agreed upon pattern will give a remainder of zero.
We see this is so in the calculation at the right.
Real applications use longer polynomials to cover larger information strings.
Some of the standard poly-nomials are listed in the text.

36. Hamming codes
The minimum Hamming distance for a code, D(min), determines its error detecting and error correcting capability.
For any code word, X, to be interpreted as a different valid code word, Y, at least D(min) single-bit errors must occur in X.
Thus, to detect k (or fewer) single-bit errors, the code must have a Hamming distance of D(min) = k + 1.

37. Flow Control and Error Control
A set of procedures that tells the sender how much data can be sent before waiting for acknowledgment
Error control
Includes both error detection and correction
Allows receiver to inform sender of lost or duplicate frames
Mostly based on Automatic Repeat Request (ARQ)

38. Data Link Protocols

39. Protocols for Noiseless Channel
Assuming channel is error free
Not realistic…
No need for error control

40. "Simplest" Mechanism Assuming Noiseless channel
Unlimited buffer and speed for the receiver

41. "Simplest" : Pseudo Code
Sender
Receiver

42. "Simplest": Flow Diagram

43. Stop-and-Wait Mechanism
Still noiseless channel
Receiver has limited buffer
Requires flow control
Sender sends one frame at a time and wait for an acknowledgment

44. Stop-and-Wait: Overview

45. Stop-and-Wait: Pseudo Code
Sender side

46. Stop-and-Wait: Pseudo Code
Receiver side

47. Stop-and-Wait: Flow Diagram

48. Noisy Channel
Realistic
Error can and will happen
Require error control
Mechanisms:
Stop-and-Wait ARQ
Go-Back-N ARQ
Selective Repeat ARQ

49. Stop-and-Wait ARQ
Sender keeps a copy of sent frame until successful delivery is ensured
Receiver responds with an ack when it successfully receives a frame
Both data and ack frames must be numbered.
(For identification purpose, both data frames and ACK frames are numbered alternately 0 and 1.)
When sender does not receive an ack within certain time, it assumes frame is lost, then retransmits the same frame.

50. Stop-and-Wait ARQ

51. Flow Diagram: Normal Operation
Sender
Receiver
S = 0
R = 0
Frame 0
Deliver
ACK 1
R = 1
S = 1
Frame 1
Deliver
ACK 0
R = 0
Time
Time

52.  Thinking Corner
Why data frames need to be numbered?

53. Flow Diagram: Lost Frame
Sender
Receiver
S = 0
R = 0
Frame 0
Deliver
ACK 1
R = 1
S = 1
Frame 1
Timeout
Frame 1
Deliver
ACK 0
R = 0
Time
Time

54. Frame 0 expected; discard
Flow Diagram: Lost ACK
Sender
Receiver
S = 0
R = 0
Frame 0
Deliver
ACK 1
R = 1
S = 1
Frame 1
Timeout
Deliver
ACK 0
R = 0
S = 1
Frame 1
Frame 0 expected; discard
ACK 0
R = 0
S = 0
Time
Time

55. Flow Diagram: Delayed ACK
Sender
Receiver
S = 0
R = 0
Frame 0
Timeout
Deliver
ACK 1
R = 1
Frame 0
Frame 0 expected; discard
S = 1
ACK 1
Frame 1
Timeout
R = 1
Frame 1
Deliver
ACK 0
R = 0
S = 0

56. Bidirectional Transmission
Data are transferred both ways
ACK are "piggybacked" with data frames

57. Improving Link Utilization
Previous example demonstrates major disadvantage of Stop-and-Wait ARQ
Prefer to send more frames before waiting for ACK
Example:
Recalculate the link utilization if we allow up to 15 frames to be sent before waiting for an ACK

58. Go-Back-N ARQ Allows multiple frames to be sent before waiting for ACK
These frames must be numbered differently
Frame numbers are called Sequence numbers
Frames must be received in the correct order
If a frame is lost, the lost frame and all of the following frames must be retransmitted

59. Sequence Numbers Frame header contains m bits for sequence number
That allows up to 2m different frame numbers
How big should m be?

60. Sending Window
Sending more than one frame at once requires sender to buffer multiple frames
Known as "sending window"
Any of these frames in the window can be lost

61. "Sliding" Window Once the first frames in the window is ACKed
ACKed frames are removed from the buffer
More frames are transmitted
Result: The window slides to the right

62. Receiving Window
Receiver expects one frame at a time

63. Send vs. Receive Windows

64. Go-Back-N: Window Sizes
For m-bit sequence numbers
Send window size: at most 2m-1
 Up to 2m-1 frames can be sent without ACK
Receive window size: 1
 Frames must be received in order

65. Go-Back-N: Normal Operation

66. Go-Back-N: Lost Frame
ACKs are cumulative

67. Lost ACK: Window Size < 2m

68. Lost ACK: Window Size = 2m

69. Selective Repeat ARQ Go-Back-N always discards out-of-order frames
Losing one frame may result in retransmission of multiple frames
Very inefficient in noisy link
Selective Repeat ARQ allows frames to be received out of order
Therefore, receive window > 1

70. Send and Receive Windows
Sender and receiver share window space equally
For m-bit sequence numbers
Send window: up to 2m-1
Receive window: up to 2m-1

71. Send Window

72. Receive Window

73. Negative ACK
Used by receiver to indicate missing frame

74. Selective Repeat: Window Size