Download Document

Computer Networks (2IC10)
Communication Protocols
Igor Radovanović
Thanks to
P. van der Stok
A. B. Forouzan
A. Tanenbaum
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Packet transmission in ideal network
1
2
2
1
3
4
5
6
t
1
2
d1
4
5
d3
d2
1
•
•
•
3
2
All packets delivered to the destination node without loss or distortion
All packets delivered in the same order
Delivery with minimum delay (d1=d2=d3)
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
6
t
N. Olifer, V. Olifer
Computer Networks,
principles, technologies and
protocols for network design,
Wiley & Sons
Communication problems
•
•
•
•
•
corruption of packets
loss of packets
replication of packets
ordering of packets
independent failures of nodes
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Packet transmission in a real network
1
2
1
3
4
5
6
t
1
2
d1
3
6
d4
d2
1’
•
•
•
•
4
d3
t
N. Olifer, V. Olifer
Computer Networks,
principles, technologies and
protocols for network design,
Wiley & Sons
Packets are delivered with variable delays (d1≠d2 ≠d3)
Packets are delivered out of order
Packets may be lost or corrupted
Average rate of information flow is different at the input and at the output of the network
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Packet corruption
• Error detection
• Error correction
Types of errors:
– single bit; only 1 bit in the data unit has changed
less likely
– burst error; 2 or more bits in the data unit have changed
Example:
0.01 s of burst impulse noise
1200 bps data rate
12 bits
of errors
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Error detection
• Redundancy: adding extra bits for detecting errors at
the destination
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Error detection
• All methods use redundancy
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Parity check
ASCII a
even
parity
• A parity bit is added to every data unit so that the total number of
1’s is even (or odd for odd parity)
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Parity check
-example• Suppose the sender wants to send the word world.
In ASCII, the five characters are coded as:
1110111 1101111 1110010 1101100 1100100
• The following shows the actual bits sent
11101110 11011110 11100100 11011000 11001001
• Parity check can detect single-bit errors
• What about burst errors?
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Error detection
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
CRC check
• based on binary division
• CRC added to the data unit so that the resulting data becomes
exactly divisible by a predetermined number
• CRC
– exactly 1 less bit than the divisor
1. A string of n zeros added to the data unit
2. New data divided by divisor
3. Replace 0’s by CRC
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
CRC generator
• Modulo-2 division
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
CRC generator (cnt’d)
5/25/2017
Name
Polynomial
Application
CRC-8
x8 + x 2 + x + 1
ATM header
CRC-10
x10 + x9 + x5 + x4 + x 2 + 1
ATM AAL
ITU-16
x16 + x12 + x5 + 1
HDLC
ITU-32
x32 + x26 + x23 + x22 + x16 + x12 + x11 + x10 + x8 +
x7 + x5 + x4 + x2 + x + 1
LANs
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
CRC performance
•
CRC can detect
1. all burst errors that affect odd number of bits
2. all burst errors of length  degree of polynomial
3. with high probability burst errors > degree of polynomial
Example:
CRC-12 (degree of 12)
1. OK
2. OK
3. 99.97 %
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Error detection
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Checksum
Example:
IP header
1.
2.
3.
4.
The data unit is divided into k sections, each of n bits
All sections are added using 1’s complement
The sum is complemented
The checksum is sent with data
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
1’s complement
• Suppose the following block of 16 bits is to be sent using a
checksum of 8 bits.
10101001 00111001
• The numbers are added using one’s complement
10101001
00111001
-----------Sum
11100010
Checksum
00011101
The pattern sent is
5/25/2017
10101001 00111001 00011101
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Checksum (cnt’d)
• Performance
– Can we use this technique to detect an error if 2 corresponding
opposite bits in 2 data segments become corrupted?
– An example:
Transmitted sequence: 10101001 00111001 00011101
Received sequence:
00101001 10111001 00011101
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Error correction
1. By retransmission
– flow and error control protocols
2. Forward Error Correction (FEC)
– require more redundancy bits
– should locate the invalid bit or bits
– n-bit code word contains m data bits + r redundancy bits
n=m+r
– m+r+1 bits discoverable by r bits
2r  m  r  1
Example:
– For m=5 data bits can r be 3?
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Hamming code
Example: 7-bit ASCII code + 4 bit redundancy
– Each r bit is the parity bit for one combination of data bits
20 1
r1: bits: 1, 3, 5, 7, 9, 11
r2: bits: 2, 3, 6, 7, 10, 11
r3: bits 4, 5, 6, 7
r4: bits 8, 9, 10, 11
5/25/2017
21
22
1
1
2 2
1
1
2 2
1
2
2
8 8
8
8
        


1 2 3 4 5 6 7 8 9 10
11
4 4 4 4
23
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Hamming code (cnt’d)
Example of redundancy bit calculation (even parity)
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Hamming code (cnt’d)
Error detection using Hamming code
once the bit is identified the receiver can reverse its value
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Communication protocols
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Communication protocols
-Wanted behavior• Communicating sender and receiver:
• Order
– When sender has sent mi before mj and receiver has accepted
mj, then receiver has also accepted mi
• Validity
– When receiver has accepted mi then mi was sent by sender
• Progress
– A sent message is eventually accepted
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Stop-and-Wait Automatic Repeat reQuest
• simplest flow and error control mechanism
• the sending device keeps a copy of the last frame
transmitted until it receives an acknowledgement
– identification of duplicate transmission (lost or delayed ACK)
• a damaged or lost frame is treated in the same way
• timers introduced
• positive ACK sent only for frames received safe &
sound
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
A Simplex Stop-and-Wait ARQ
1.
2.
3.
4.
normal operation
the frame is lost
the ACK is lost
the ACK is delayed
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Stop-and-Wait ARQ
-Normal operation• The sender will not send the
next piece of data until it is
sure that the current one is
correctly received
• sequence number necessary
to check for duplicated
packets
• No NACK – when packet is
corrupted – duplicate ACKs
instead
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Stop-and-Wait ARQ
-Lost or damaged frame-
roundtrip delay
+
processing in the
receiver
Should be as short
as possible. Why?
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Stop-and-Wait ARQ
-Lost ACK-
Importance of
numbering
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Stop-and-Wait ARQ
-Delayed ACK-
Importance of
ACK numbering
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Duplex Stop-and-Wait ARQ
• Piggybacking
– combine data with ACK (less overhead saves bandwidth)
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Drawbacks of Stop-and-Wait ARQ
•
after each frame sent the host must wait for an ACK
– inefficient use of bandwidth
•
to improve efficiency ACK should be sent after multiple frames
•
Alternatives: Sliding Window protocols
1. Go-back-N ARQ
2. Selective Repeat ARQ
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Pipelining
• One task begins before the other one ends
– increases efficiency in transmission
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Sliding Window Protocols
• Sequence numbers
– sent frames are numbered sequentially
– number of frames stored in the header
• if the number of bits in the header is m than
sequence number goes from 0 to 2^m-1
• Sliding window
frame
– to hold the
unacknowledged
outstanding frames
– the receiver
window size
always 1
acknowledged
frames
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
sequence
number
Question
• What is the consequence of the window size in the
receiver on the order of the received packets?
Answer:
• Window size 1 means that the packets are received in
order
– Note: this is not the case for the larger receiver window size
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Go-back-N
-Control variables•
•
•
•
S- holds the sequence number of the recently sent frame
SF – holds sequence number of the first frame in the window
SL – holds the sequence number of the last frame
R – sequence number of the frame expected to be received
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
The name of Go-back-N: why?
• Re-sending frame
– when the frame is damaged the sender goes back and sends
a set of frames starting from the last one ACKn’d
– the number of retransmitted frames is N
Example:
The window size is 4.
A sender has sent frame 6 and the timer expires for
frame 3 (frame 3 not ACKn’d). The sender goes back
and re-sends the frames 3, 4, 5 and 6.
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Go-back-N
-normal operation• How many frames
can be transmitted
without
acknowledgment?
• ACK1 – not
necessary if ACK2
is sent
– Cumulative ACK
expected sequence
number
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Go-back-N
-damaged or lost framedamaged frames
are discarded!
Why are correctly
received packets
not buffered?
What is the
disadvantage of
this?
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Go-back-N
-sender window size-
sequence
number
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Go-back-N
• Inefficient
– all out of order received packets are discarded
• This is a problem in a noisy link
– many frames must be retransmitted -> bandwidth consuming
• Solution
– re-send only the damaged frames
• Selective Repeat ARQ
– avoid unnecessary retransmissions
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Selective Repeat ARQ
• Processing at the receiver more complex
• The window size is reduced to 2m-1 (at most)
• Both the transmitter and the receiver have the same
window size
• Receiver expects frames within the range of the
sequence numbers
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Selective Repeat ARQ
-lost frame-
Note:
retransmission
triggered with NACK
and not with
expired timer
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Selective Repeat ARQ
-sender window size-
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Communication problems
•
•
•
•
•
corruption of packets
loss of packets
replication of packets
ordering of packets
independent failures of nodes
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Communication protocols
• We have assumed until now that packets cannot be
reordered
– this when the parties are connected by a single physical line
• What if the channel is a network?
– the packets may arrive out of order
• Sequence number might be reused – duplications
possible
• How to solve this problem?
• Introduce the lifetime of the packet
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Connection establishment
• Connection request PDU
• Connection accepted PDU
• Is this enough?
– What if the network can store or duplicate packets?
– What if the subnet is so congested that ACK or datagrams
cannot go through?
Example: bank transaction (delayed duplicates)
Solutions:
– new addresses for each connection?
– incremented sequence numbers?
– PACKET LIFETIME
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Packet lifetime
• Can be restricted using:
– restricted subnet design
• prevents looping
• bounding congestion delay over the longest possible path
– putting a hop counter in each packet
• decrementing default value & dropping when zero
– time stamping each packet
• carry the time when it was created
• discarded after some predefined time
– all the routers synchronized – nontrivial
• Lifetime expires:
– Not only the packets but all the ACKs must be dead
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Connection Establishment (2)
dead time
after crash
data rate < clk rate
(max packet
lifetime)
the same seq num
connection 5
crash
restart
Time-of-day clock (not necessarily sync) – 2 PDU with the same
seq. num. are never outstanding at the same time
solves the delayed duplicate problem for data PDU
(a) PDUs may not enter the forbidden region.
(b) The resynchronization problem.
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Connection Establishment (2)
different sequence numbers
used by the sender & receiver
Three protocol scenarios for establishing a connection using a three-way
handshake. CR denotes CONNECTION REQUEST.
(a) Normal operation,
(b) Old CONNECTION REQUEST appearing out of nowhere.
(c) Duplicate CONNECTION REQUEST and duplicate ACK.
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Connection Release
• asymmetric release
– a telephone line
• connection is broken when
one party hangs up
– abrupt disconnection with loss of
data.
• symmetric release
– each connection released
separately
– continue to receive after
disconnect PDU is sent
– avoid data loss
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Connection Release (2)
unreliable communication
channel
The two-army problem.
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Connection Release (3)
disconnect request
6-14, a, b
Four protocol scenarios for releasing a connection.
(a) Normal case of a three-way handshake. (b) final ACK lost.
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking
Connection Release (4)
6-14, c,d
(c) Response lost. (d) Response lost and subsequent DRs lost.
5/25/2017
Igor Radovanović, [email protected]
TU/e Computer Science, System Architecture and Networking