flow control and error control

Flow control and Error control

ECS 152AXin 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: IntroductionSome 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

“link”

data-link layer has responsibility of transferring datagram from one node to adjacent node over a 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

link layer implemented in “adaptor” (aka NIC) Ethernet card, PCMCI card,

802.11 card

sending side: encapsulates datagram in

a frame adds error checking bits,

rdt, flow control, etc.

receiving side looks for errors, rdt, flow

control, etc extracts datagram,

passes to rcving node

adapter is semi-autonomous

link & physical layers

sendingnode

frame

rcvingnode

datagram

frame

adapter adapter

link layer protocol

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 DetectionEDC= 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

Sender: treat segment contents

as sequence of 16-bit integers

checksum: addition (1’s complement sum) of segment contents

sender puts checksum value into UDP checksum field

Receiver: compute checksum of

received segment check if computed checksum

equals checksum field value: NO - error detected YES - no error detected.

But maybe errors nonetheless? More later ….

Goal: detect “errors” (e.g., flipped bits) in transmitted segment

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 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1

1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1

1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1

wraparound

sumchecksum

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 ExampleWant:

D.2r XOR R = nGequivalently:

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

(x7 x6 1) (x6 x5 ) x7 (1 1)x6 x 5 1

x7 x5 1

(x 1)(x2 x 1) x3 x 2 x x2 x 1 x3 1

Addition:

Multiplication:

Division: x3 + x + 1 ) x6 + x5

x3 + x2 + x

x6 + x4 + x3

x5 + x4 + x3

x5 + x3 + x2

x4 + x2

x4 + x2 + xx

= q(x) quotient

= r(x) remainder

divisordividend

35 ) 1223

10517

Figure 3.55

Cyclic Redunancy 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)

3) Add remainder r(x) to xn-k i(x) (puts check bits in the n-k low order positions):

quotient remainder

transmitted codewordb(x) = xn-ki(x) + r(x)

xn-ki(x) = g(x) q(x) + r(x)

Cyclic Redundancy Check

Cyclic Redundancy Check

Generator polynomial: g(x)= x3 + x + 1Information: (1,1,0,0) i(x) = x3 + x2

Encoding: x3i(x) = x6 + x5

1011 ) 1100000

1110

10111110101110101011

010

x3 + x + 1 ) x6 + x5

x3 + x2 + x

x6 + x4 + x3

x5 + x4 + x3

x5 + x3 + x2

x4 + x2

x4 + x2 + xx

Transmitted codeword:b(x) = x6 + x5 + xb = (1,1,0,0,0,1,0)

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 00000

01 00111

10 11001

11 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

data from above : if next available seq # in

window, send pkt

timeout(n): 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 #

senderpkt n in [rcvbase, rcvbase+N-

1]

send ACK(n) out-of-order: buffer in-order: deliver (also

deliver buffered, in-order pkts), advance window to next not-yet-received pkt

pkt n in [rcvbase-N,rcvbase-1]

ACK(n)

otherwise: ignore

receiver

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?