Flow control and Error control
ECS 152A
Xin Liu
Based on Kurose and Ross
The Data Link Layer
Our goals:
 understand principles behind data link layer
services:



Flow control
error detection, correction
reliable data transfer
Link Layer: Introduction
Some terminology:
 hosts and routers are nodes
 communication channels that
connect adjacent nodes along
communication path are links



wired links
wireless links
LANs
 layer-2 packet is a frame,
encapsulates datagram
data-link layer has responsibility of
transferring datagram from one node
to adjacent node over a link
“link”
Link layer: context
 Datagram transferred by
different link protocols
over different links:

e.g., Ethernet on first link,
frame relay on
intermediate links, 802.11
on last link
 Each link protocol
provides different
services

e.g., may or may not
provide rdt over link
transportation analogy
 trip from Princeton to
Lausanne
 limo: Princeton to JFK
 plane: JFK to Geneva
 train: Geneva to Lausanne
 tourist = datagram
 transport segment =
communication link
 transportation mode =
link layer protocol
 travel agent = routing
algorithm
Link Layer Services
 Framing, link access:



encapsulate datagram into frame, adding header, trailer
channel access if shared medium
“MAC” addresses used in frame headers to identify
source, dest
• different from IP address!
 Reliable delivery between adjacent nodes
 we learned how to do this already (chapter 3)!
 seldom used on low bit error link (fiber, some twisted
pair)
 wireless links: high error rates
• Q: why both link-level and end-end reliability?
Link Layer Services (more)
 Flow Control:

pacing between adjacent sending and receiving nodes
 Error Detection:


errors caused by signal attenuation, noise.
receiver detects presence of errors:
• signals sender for retransmission or drops frame
 Error Correction:

receiver identifies and corrects bit error(s) without
resorting to retransmission
 Half-duplex and full-duplex
 with half duplex, nodes at both ends of link can transmit,
but not at same time
Adaptors Communicating
datagram
sending
node
frame
adapter
rcving
node
link layer protocol
frame
adapter
 link layer implemented in  receiving side
“adaptor” (aka NIC)
 looks for errors, rdt, flow
control, etc
 Ethernet card, PCMCI
 extracts datagram, passes
card, 802.11 card
to rcving node
 sending side:
 adapter is semi encapsulates datagram in
autonomous
a frame
 adds error checking bits,
 link & physical layers
rdt, flow control, etc.
Flow Control
 Ensuring the sending entity does not
overwhelm the receiving entity

Preventing buffer overflow
 Transmission time
 Time
taken to emit all bits into medium
 Propagation time

Time for a bit to traverse the link
Model of Frame Transmission
Stop and Wait
 Source transmits frame
 Destination receives frame and replies with
acknowledgement
 Source waits for ACK before sending next
frame
 Destination can stop flow by not send ACK
 Works well for a few large frames
Fragmentation
 Large block of data may be split into small
frames
Limited buffer size
 Errors detected sooner (when whole frame
received)
 On error, retransmission of smaller frames is
needed
 Prevents one station occupying medium for long
periods

 Stop and wait becomes inadequate
Stop and Wait Link Utilization
Sliding Windows Flow Control
 Allow multiple frames to be in transit
 Receiver has buffer W long
 Transmitter can send up to W frames without ACK
 Each frame is numbered
 ACK includes number of next frame expected
 Sequence number bounded by size of field (k)
 Frames are numbered modulo 2k
Sliding Window Diagram
Example Sliding Window
Sliding Window Enhancements
 Receiver can acknowledge frames without
permitting further transmission (Receive Not
Ready)
 Must send a normal acknowledge to resume
 If duplex, use piggybacking


If no data to send, use acknowledgement frame
If data but no acknowledgement to send, send last
acknowledgement number again, or have ACK valid flag
(TCP)
Q: if the size of field is k, what is the maximum window size?
A: 2^k -1. Assuming 0,1,…2^k-1 (2^k window size) sent out,
need to ack the last 2^k-1.
Error Detection and correction
Error Detection
EDC= Error Detection and Correction bits (redundancy)
D = Data protected by error checking, may include header fields
• Error detection not 100% reliable!
• protocol may miss some errors, but rarely
• larger EDC field yields better detection and correction
Parity Checking
Single Bit Parity:
Detect single bit errors
Two Dimensional Bit Parity:
Detect and correct single bit errors
0
0
UDP checksum
Goal: detect “errors” (e.g., flipped bits) in transmitted
segment
Sender:
Receiver:
 treat segment contents
 compute checksum of
as sequence of 16-bit
integers
 checksum: addition (1’s
complement sum) of
segment contents
 sender puts checksum
value into UDP checksum
field
received segment
 check if computed checksum
equals checksum field value:
 NO - error detected
 YES - no error detected.
But maybe errors
nonetheless? More later
….
Internet Checksum Example
 Note

When adding numbers, a carryout from the
most significant bit needs to be added to the
result
 Example: add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0
1 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
wraparound 1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
sum 1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 0
checksum 1 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
Cyclic Redundancy Check
 view data bits, D, as a binary number
 choose r+1 bit pattern (generator), G
 goal: choose r CRC bits, R, such that



<D,R> exactly divisible by G (modulo 2)
receiver knows G, divides <D,R> by G. If non-zero remainder:
error detected!
can detect all burst errors less than r+1 bits
 widely used in practice (ATM, HDCL)
CRC Example
Want:
D.2r XOR R = nG
equivalently:
D.2r = nG XOR R
equivalently:
if we divide D.2r by
G, want remainder R
R = remainder[
D.2r
G
]
Cyclic Redundancy Check
 For a block of k bits transmitter generates
n bit sequence
 Transmit k+n bits which is exactly divisible
by some number
 Receive divides frame by that number

If no remainder, assume no error
Cyclic Redunancy check
Addition:
(x 7  x 6  1)  (x 6  x 5 )  x 7  (1  1)x 6  x 5  1
 x7  x5  1
Multiplication:
Division:
divisor
3
35 ) 122
105
17
(x  1)(x  x  1)  x  x  x  x  x  1  x  1
2
3
2
2
3
= q(x) quotient
x3 + x2 + x
x3 + x + 1 ) x6 + x5
x6 +
x4 + x3
dividend
x5 + x4 + x3
x5 +
x3 + x2
x4 +
x4 +
x2
x2 + x
x
= r(x) remainder
Figure 3.55
Cyclic Redundancy Check
Steps:
1) Multiply i(x) by xn-k (puts zeros in (n-k) low order positions)
2) Divide xn-k i(x) by g(x)
quotient
remainder
xn-ki(x) = g(x) q(x) + r(x)
3) Add remainder r(x) to xn-k i(x)
(puts check bits in the n-k low order positions):
b(x) = xn-ki(x) + r(x)
transmitted codeword
Cyclic Redundancy Check
Generator polynomial: g(x)= x3 + x + 1
Information: (1,1,0,0)
i(x) = x3 + x2
Encoding: x3i(x) = x6 + x5
x3 + x2 + x
x3 + x + 1 ) x6 + x5
x6 +
x 4 + x3
x5 + x4 + x3
x5 +
x 3 + x2
x4 +
x2
x4 +
x2 + x
x
Transmitted codeword:
b(x) = x6 + x5 + x
b = (1,1,0,0,0,1,0)
1110
1011 ) 1100000
1011
1110
1011
1010
1011
010
Capability of CRC
 An error E(X) is undetectable if it is divisible by
P(x). The following can be detected.






All single-bit errors if P(x) has more than one nonzero
term
All double-bit errors if P(x) has a factor with three
terms
Any odd number of errors, if P(x) contain a factor x+1
Any burst with length less or equal to n-k
A fraction of error burst of length n-k+1; the fraction is
1-2^(-(-n-k-1)).
A fraction of error burst of length greater than n-k+1;
the fraction is 1-2^(-(n-k)).
 Powerful error detection; more computation
complexity compared to Internet checksum
Error Correction
 Correction of detected errors usually requires
data block to be retransmitted Not appropriate
for wireless applications

Bit error rate is high
• Lots of retransmissions

Propagation delay can be long (satellite) compared with
frame transmission time
• Would result in retransmission of frame in error plus
many subsequent frames
 Need to correct errors on basis of bits received
 Basic idea: redundancy
Error Correction
 Hamming distance: d(v1,v2) between two
n-bit binary sequences v1 and v2 is the #
of bits in which v1 and v2 disagree.
 (n,k) block code: map each k-bit sequence
into a unique n-bit codeword (n>k).
 K=2,n=5
Data
code
00
01
00000
00111
10
11
11001
11110
Consider one bit of error, say
00000->00001.
d(00000,00001)=1;
d(00111,00001)=2;
d(11001,00001)=2;
d(11110,00001)=4;
Error Control
 Detection and correction of errors
 Lost frames
 Damaged frames
 Automatic repeat request
Error detection
 Positive acknowledgement
 Retransmission after timeout
 Negative acknowledgement

Automatic Repeat Request
(ARQ)
 Stop and wait
 Go back N
 Selective reject (selective retransmission)
Stop and Wait
 Source transmits single frame
 Wait for ACK
 If received frame damaged, discard it
 Transmitter has timeout
 If no ACK within timeout, retransmit
 If ACK damaged,transmitter will not
recognize it
 Transmitter
will retransmit
 Receive gets two copies of frame
 Use ACK0 and ACK1
Stop and Wait Diagram
Stop and Wait - Pros and Cons
 Simple
 Inefficient
Go Back N (1)
 Based on sliding window
 If no error, ACK as usual with next frame
expected
 Use window to control number of
outstanding frames
 If error, reply with rejection
Discard that frame and all future frames until
error frame received correctly
 Transmitter must go back and retransmit that
frame and all subsequent frames

Go Back N - Damaged Frame
 Receiver detects error in frame i
 Receiver sends rejection-i
 Transmitter gets rejection-i
 Transmitter retransmits frame i and all
subsequent
Go Back N - Lost Frame (1)
 Frame i lost
 Transmitter sends i+1
 Receiver gets frame i+1 out of sequence
 Receiver send reject i
 Transmitter goes back to frame i and
retransmits
Go Back N - Lost Frame (2)
 Frame i lost and no additional frame sent
 Receiver gets nothing and returns neither
acknowledgement nor rejection
 Transmitter times out and sends
acknowledgement frame with P bit set to 1
 Receiver interprets this as command which
it acknowledges with the number of the
next frame it expects (frame i )
 Transmitter then retransmits frame i
Go Back N - Damaged
Acknowledgement
 Receiver gets frame i and send
acknowledgement (i+1) which is lost
 Acknowledgements are cumulative, so next
acknowledgement (i+n) may arrive before
transmitter times out on frame i
 If transmitter times out, it sends
acknowledgement with P bit set as before
 This can be repeated a number of times
before a reset procedure is initiated
Go Back N - Damaged Rejection
 As for lost frame (2)
Go Back N Diagram
Selective Repeat
 receiver individually acknowledges all correctly
received pkts

buffers pkts, as needed, for eventual in-order delivery
to upper layer
 sender only resends pkts for which ACK not
received

sender timer for each unACKed pkt
 sender window
 N consecutive seq #’s
 again limits seq #s of sent, unACKed pkts
Selective repeat: sender, receiver windows
Selective repeat
sender
data from above :
receiver
pkt n in [rcvbase, rcvbase+N-1]
 if next available seq # in
 send ACK(n)
timeout(n):
 in-order: deliver (also
window, send pkt
 resend pkt n, restart timer
ACK(n) in [sendbase,sendbase+N]:
 mark pkt n as received
 if n smallest unACKed pkt,
advance window base to
next unACKed seq #
 out-of-order: buffer
deliver buffered, in-order
pkts), advance window to
next not-yet-received pkt
pkt n in
[rcvbase-N,rcvbase-1]
 ACK(n)
otherwise:
 ignore
Selective repeat in action
Selective repeat:
dilemma
Example:
 seq #’s: 0, 1, 2, 3
 window size=3
 receiver sees no
difference in two
scenarios!
 incorrectly passes
duplicate data as new
in (a)
Q: what relationship
between seq # size
and window size?