Transport Layer 3-1
Chapter 3Transport Layer
Computer Networking: A Top Down Approach
6th edition Jim Kurose, Keith Ross
Addison-WesleyMarch 2012
All material copyright 1996-2012J.F Kurose and K.W. Ross, All Rights Reserved
蒋力计算机科学与工程系
Transport Layer 3-2
Chapter 3: Transport Layer
our goals: understand
principles behind transport layer services: multiplexing,
demultiplexing
reliable data transfer
flow control
congestion control
learn about Internet transport layer protocols: UDP: connectionless
transport
TCP: connection-oriented reliable transport
TCP congestion control
Transport Layer 3-3
Chapter 3 outline
3.1 transport-layer services
3.2 multiplexing and demultiplexing
3.3 connectionless transport: UDP
3.4 principles of reliable data transfer
3.5 connection-oriented transport: TCP
segment structure
reliable data transfer
flow control
connection management
3.6 principles of congestion control
3.7 TCP congestion control
Transport Layer 3-4
Transport services and protocols
provide logical communicationbetween app processes running on different hosts
transport protocols run in end systems
send side: breaks app messages into segments, passes to network layer
rcv side: reassembles segments into messages, passes to app layer
Router won’t check segments
more than one transport protocol available to apps
Internet: TCP and UDP
applicationtransportnetworkdata linkphysical
applicationtransportnetworkdata linkphysical
Transport Layer 3-53-5
Transport vs. network layer
network layer: logical communication between hosts
transport layer:logical communication between processes relies on, enhances,
network layer services
12 kids in Ann’s house sending letters to 12 kids in Bill’s house:
hosts = houses processes = kids app messages = letters in
envelopes transport protocol = Ann
and Bill who demux to in-house siblings
network-layer protocol = postal service
household analogy:
Ann’s siblings regard Ann as postal service, but Ann works in house not in real postal service.
When Ann and Bill travel, Susan and Harvey take responsibility. They are younger and unreliable!
Ann and Bill’s service rely on postal service! But they can help a little! Lost, security etc.
Transport Layer 3-6
Internet transport-layer protocols
reliable, in-order delivery (TCP) congestion control
flow control
connection setup
unreliable, unordered delivery: UDP no-frills extension of “best-
effort” IP Minimum service: deliver,
error-checking
services not available: delay guarantees
bandwidth guarantees
applicationtransportnetworkdata linkphysical
applicationtransportnetworkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
Network links overflow
Receiver overflow
Transport Layer 3-73-7
Chapter 3 outline
3.1 transport-layer services
3.2 multiplexing and demultiplexing
3.3 connectionless transport: UDP
3.4 principles of reliable data transfer
3.5 connection-oriented transport: TCP
segment structure
reliable data transfer
flow control
connection management
3.6 principles of congestion control
3.7 TCP congestion control
Transport Layer 3-8
Multiplexing/demultiplexing
process
socket
use header info to deliverreceived segments to correct socket
demultiplexing at receiver:multiplexing at sender:
handle data from multiplesockets, add transport header (later used for demultiplexing)
transport
application
physical
link
network
P2P1
transport
application
physical
link
network
P4
transport
application
physical
link
network
P3
Rely on sockets
Multiplexing/demultiplexing is not limited in transport level. Their use depends on when a single protocol in a level are invoked by higher level protocol
Ann and Bill
Transport Layer 3-9
How demultiplexing works
host receives IP datagrams each datagram has source IP
address, destination IP address
each datagram carries one transport-layer segment
each segment has source, destination port number
host uses IP addresses & port numbers to direct segment to appropriate socket
source port # dest port #
32 bits
applicationdata
(payload)
other header fields
TCP/UDP segment format
Transport Layer 3-10
Connectionless demultiplexing
recall: created socket has host-local port #:clientSocket = socket(AF_INET,SOCK_DGRAM)
when host receives UDP segment: checks destination port #
in segment
directs UDP segment to socket with that port #
recall: when creating datagram to send into UDP socket, must specify destination IP address
destination port #
IP datagrams with same dest. port #, but different source IP addresses and/or source port numbers will be directed to same socket at dest
OS assign one for you
clientSocket.bind((‘’,12534))
clientSocket.sendto(message,(serverIP,serverPort))
serverSocket.recvfrom(serverPort)
Transport Layer 3-11
Connectionless demux: exampleserverSocket = socket(AF_INET, SOCK_DGRAM)
serverSocket.bind((‘’, 6428));
transport
application
physical
link
network
P3transport
application
physical
link
network
P1
transport
application
physical
link
network
P4
clientSocket1 = socket(AF_INET, SOCK_DGRAM)
clientSocket.bind((‘’, 5775));
clientSocket2 = socket(AF_INET, SOCK_DGRAM)
clientSocket.bind((‘’, 9157));
source port: 9157dest port: 6428
source port: 6428dest port: 9157
source port: ?dest port: ?
source port: ?dest port: ?
Transport Layer 3-12
Connection-oriented demux
TCP socket identified by 4-tuple: source IP address
source port number
dest IP address
dest port number
demux: receiver uses all four values to direct segment to appropriate socket
server host may support many simultaneous TCP sockets: each socket identified by
its own 4-tuple
web servers have different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 2-13
实例程序: TCP client
from socket import *
serverName = ’servername’
serverPort = 12000
clientSocket = socket(AF_INET, SOCK_STREAM)
clientSocket.connect((serverName,serverPort))
sentence = raw_input(‘Input lowercase sentence:’)
clientSocket.send(sentence)
modifiedSentence = clientSocket.recv(1024)
print ‘From Server:’, modifiedSentence
clientSocket.close()
Python TCPClient
为接触服务器创建TCP 套接字, 远程端口号12000
不需要附加地址和端口号
Transport Layer 2-14
示例程序: TCP server
from socket import *serverPort = 12000
serverSocket = socket(AF_INET,SOCK_STREAM)serverSocket.bind((‘’,serverPort))
serverSocket.listen(1)print ‘The server is ready to receive’
while 1:connectionSocket, addr = serverSocket.accept()
sentence = connectionSocket.recv(1024)capitalizedSentence = sentence.upper()
connectionSocket.send(capitalizedSentence)connectionSocket.close()
Python TCPServer
创建TCP 欢迎套接字
服务器开始监听到来的TCP 请求
循环
服务器在accept()等待到达的请求, 新的套接字被创
建
从套接字读取字节流(但不像UDP那样读地址)
对这个客户关闭套接字 (但不关闭欢迎套接字)
Transport Layer 3-15
Connection-oriented demux: example
transport
application
physical
link
network
P3transport
application
physical
link
P4
transport
application
physical
link
network
P2
source IP,port: A,9157dest IP, port: B,80
source IP,port: B,80dest IP,port: A,9157
host: IP address A
host: IP address C
network
P6P5P3
source IP,port: C,5775dest IP,port: B,80
source IP,port: C,9157dest IP,port: B,80
three segments, all destined to IP address: B,dest port: 80 are demultiplexed to different sockets
server: IP address B
Transport Layer 3-16
Connection-oriented demux: example
transport
application
physical
link
network
P3transport
application
physical
link
transport
application
physical
link
network
P2
source IP,port: A,9157dest IP, port: B,80
source IP,port: B,80dest IP,port: A,9157
host: IP address A
host: IP address C
server: IP address B
network
P3
source IP,port: C,5775dest IP,port: B,80
source IP,port: C,9157dest IP,port: B,80
P4
threaded server
Would it be a problem if two clients send pkt with same source port #
Review Q: Why most TCP services use a monitoring socket?
Transport Layer
Connection-oriented demux: HTTP
Check out the python implemented simple HTTP server
Port scanner
Transport Layer 3-18
Chapter 3 outline
3.1 transport-layer services
3.2 multiplexing and demultiplexing
3.3 connectionless transport: UDP
3.4 principles of reliable data transfer
3.5 connection-oriented transport: TCP
segment structure
reliable data transfer
flow control
connection management
3.6 principles of congestion control
3.7 TCP congestion control
Transport Layer 3-19
UDP: User Datagram Protocol [RFC 768]
“no frills,” “bare bones”Internet transport protocol
“best effort” service, UDP segments may be:
lost
delivered out-of-order to app
connectionless:
no handshaking between UDP sender, receiver
each UDP segment handled independently of others
UDP use: streaming multimedia
apps (loss tolerant, rate sensitive)
DNS
SNMP
reliable transfer over UDP: add reliability at
application layer
application-specific error recovery!
Transport Layer 3-20
Why is there a UDP?
Flexible for app to decide when and what to send No congestion control: UDP can blast away as fast as desired TCP will re-send pkt if out-of-order or corrupt
No connection setup time 3 RRT handshaking
No connection state at end-systems Smaller header size Media stream app - UDP or TCP?
Massive stream data congests the network
Take-home work use wireshark to investigate the video web & their app Tips: iku, open 8909, 17171, 8828 Tips: CDN
Transport Layer 3-21
UDP: segment header
source port # dest port #
32 bits
applicationdata
(payload)
UDP segment format
length checksum
length, in bytes of UDP segment,
including header
no connection establishment (which can add delay)
simple: no connection state at sender, receiver
small header size
no congestion control: UDP can blast away as fast as desired
why is there a UDP?
Transport Layer 3-22
UDP checksum
sender: treat segment contents,
including header fields, as sequence of 16-bit integers
checksum: addition (one’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
Transport Layer 3-23
Internet checksum: example
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
sum
checksum
Note: when adding numbers, a carryout from the most significant bit needs to be added to the result
Transport Layer 3-24
Why Checksum at Transport Level?
Is Link Layer Enough?
Data transferred in links are correct
No Guarantees
But they can be corrupted in router and network layer
End-to-End Principle
Transport Layer 3-25
Application View of the World
Transport Layer 3-26
Why Doesn’t the Network Help?
Compress data?
Reformat/translate/improve requests?
Serve cached data?
Add security?
Migrate connections across the network?
Or one of any of a huge number of other things?
Generally, may things the network can do to improve the app and ease the design.
But NO
Transport Layer 3-27
End-to-End Principle“The function in question can completely and correctly be implemented only with the knowledge and help of the application standing at the end
points of the communication system.”
“Functions placed at the lower levels may be redundant or of little valuewhen compared to providing them at the higher level.”
Saltzer 1984
Do we need link layer checksum? Think about Cost!!
“Communications protocol operations should be defined to occur at the end-points of a communications system.”
“Sometimes an incomplete version of the function provided by the communication system may be useful as a performance enhancement.”
Transport Layer 3-28
Example: File Transfer
Exactly some program in MIT did!
The assumption is wrong and developer lost lots of the source code
TCP error checking CANNOT guarantee HASH
Link error checking CAN help/improve the performance, e.g., wireless!
Transport Layer 3-29
“Strong” End-to-End Principle
The network’s job is to transmit datagrams as
efficiently and flexibly as possible. Everything else
should be done at the fringes…-[RFC 1958]
Transport Layer 3-303-30
Chapter 3 outline
3.1 transport-layer services
3.2 multiplexing and demultiplexing
3.3 connectionless transport: UDP
3.4 principles of reliable data transfer
3.5 connection-oriented transport: TCP
segment structure
reliable data transfer
flow control
connection management
3.6 principles of congestion control
3.7 TCP congestion control
Transport Layer 3-31
Principles of reliable data transfer important in application, transport, link layers
top-10 list of important networking topics!
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Pkt corruption, Pkt loss, Pkt duplication
Transport Layer 3-32
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Principles of reliable data transfer important in application, transport, link layers
top-10 list of important networking topics!
Transport Layer 3-33
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
important in application, transport, link layers top-10 list of important networking topics!
Principles of reliable data transfer
Transport Layer 3-34
Reliable data transfer: getting started
send
side
receive
side
rdt_send(): called from above,
(e.g., by app.). Passed data to deliver to receiver upper layer
udt_send(): called by rdt,
to transfer packet over unreliable channel to receiver
rdt_rcv(): called when packet
arrives on rcv-side of channel
deliver_data(): called by
rdt to deliver data to upper
Review Q: How many
times it calls?
Review Q: How many
times it calls?
Transport Layer 3-35
we’ll: incrementally develop sender, receiver sides of
reliable data transfer protocol (rdt)
consider only unidirectional data transfer but control info will flow on both directions!
use finite state machines (FSM) to specify sender, receiver
state1
state2
event causing state transition
actions taken on state transition
state: when in this “state” next state
uniquely determined by next event
event
actions
Reliable data transfer: getting started
Transport Layer 3-36
rdt1.0: reliable transfer over a reliable channel
underlying channel perfectly reliable no bit errors
no loss of packets
separate FSMs for sender, receiver: sender sends data into underlying channel
receiver reads data from underlying channel
Wait for
call from
above packet = make_pkt(data)
udt_send(packet)
rdt_send(data)
extract (packet,data)
deliver_data(data)
Wait for
call from
below
rdt_rcv(packet)
sender receiver
Transport Layer 3-37
underlying channel may flip bits in packet checksum to detect bit errors
the question: how to recover from errors: acknowledgements (ACKs): receiver explicitly tells sender
that pkt received OK negative acknowledgements (NAKs): receiver explicitly tells
sender that pkt had errors sender retransmits pkt on receipt of NAK
new mechanisms in rdt2.0 (beyond rdt1.0): error detection receiver feedback: control msgs (ACK,NAK) rcvr-
>sender
rdt2.0: channel with bit errors
How do humans recover from “errors”during conversation?
Let's play a game:Transmit a sentence to the on on your back!
i. the listener cannot talk
ii. the listener can talk
Transport Layer 3-38
underlying channel may flip bits in packet
How to detect errors? checksum to detect bit errors
How to recover from errors: Receiver Feedback
acknowledgements (ACKs): receiver explicitly tells sender that pkt
received OK
negative acknowledgements (NAKs): receiver explicitly tells
sender that pkt had errors;
sender retransmits pkt on receipt of NAK
rdt2.0: channel with bit errors
Transport Layer 3-39
new mechanisms in rdt2.0 (beyond rdt1.0): error detection feedback: control msgs (ACK,NAK) from receiver to
sender
rdt2.0: Discussion
Sender How to construct pkt?
Add checksum to the packet
Must check feedback If ACK?
Send next pkt
If NAK? Resend this pkt
Receiver Must check if received
packet is corrupted If not, send back ACK
Otherwise, send back NAK
Transport Layer 3-40
rdt2.0: FSM specification
Wait for
call from
above
sndpkt = make_pkt(data, checksum)
udt_send(sndpkt)
extract(rcvpkt,data)
deliver_data(data)
udt_send(ACK)
rdt_rcv(rcvpkt) &&
notcorrupt(rcvpkt)
rdt_rcv(rcvpkt) && isACK(rcvpkt)
udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
isNAK(rcvpkt)
udt_send(NAK)
rdt_rcv(rcvpkt) &&
corrupt(rcvpkt)
Wait for
ACK or
NAK
Wait for
call from
belowsender
receiverrdt_send(data)
L
Transport Layer 3-41
rdt2.0: operation with no errors
Wait for
call from
above
sndpkt = make_pkt(data, checksum)
udt_send(sndpkt)
extract(rcvpkt,data)
deliver_data(data)
udt_send(ACK)
rdt_rcv(rcvpkt) &&
notcorrupt(rcvpkt)
rdt_rcv(rcvpkt) && isACK(rcvpkt)
udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
isNAK(rcvpkt)
udt_send(NAK)
rdt_rcv(rcvpkt) &&
corrupt(rcvpkt)
Wait for
ACK or
NAK
Wait for
call from
below
rdt_send(data)
L
Transport Layer 3-42
rdt2.0: error scenario
Wait for
call from
above
snkpkt = make_pkt(data, checksum)
udt_send(sndpkt)
extract(rcvpkt,data)
deliver_data(data)
udt_send(ACK)
rdt_rcv(rcvpkt) &&
notcorrupt(rcvpkt)
rdt_rcv(rcvpkt) && isACK(rcvpkt)
udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
isNAK(rcvpkt)
udt_send(NAK)
rdt_rcv(rcvpkt) &&
corrupt(rcvpkt)
Wait for
ACK or
NAK
Wait for
call from
below
rdt_send(data)
L
Transport Layer 3-43
rdt2.0 has a fatal flaw!
what happens if ACK/NAK corrupted?
sender doesn’t know what happened at receiver!
What to do?
Retransmit?
can’t just retransmit: possible duplicate
handling duplicates: sender retransmits current pkt
if ACK/NAK corrupted
How receiver know the pkt is resent (Y/N deliver)?
sender adds sequence number to each pkt
receiver discards (doesn’t deliver up) duplicate pkt
stop and waitsender sends one packet, then waits for receiver response
Review Q: How many sequence number do we
need?
Transport Layer 3-44
rdt2.1: Exploration
Sender Seq # Added to Packet
2 seq #’s (0,1) will suffice why?
Must check if received ACK/NAK corrupted
If corrupted? resend the pkt
with the same seq #
Receiver Must check if received
packet is duplicate
State indicates whether 0 or 1 is expected pktseq#
If duplicated?NOT deliver
Must resend ACK when receiving a duplicate
Note
Receiver can not know if its last ACK/NAK was received OK at sender
Transport Layer 3-45
rdt2.1: sender, handles garbled ACK/NAKs
Wait for
call 0 from
above
rdt_send(data)
sndpkt = make_pkt(0, data, checksum)
udt_send(sndpkt)
Wait for
ACK or
NAK 0
rdt_send(data)
sndpkt = make_pkt(1, data, checksum)
udt_send(sndpkt)
rdt_rcv(rcvpkt)
&& notcorrupt(rcvpkt)
&& isACK(rcvpkt)
rdt_rcv(rcvpkt)
&& notcorrupt(rcvpkt)
&& isACK(rcvpkt)
Wait for
call 1 from
above
Wait for
ACK or
NAK 1
LL
Transport Layer 3-46
Wait for
0 from
below
rdt_rcv(rcvpkt) && notcorrupt(rcvpkt)
&& has_seq1(rcvpkt)
extract(rcvpkt,data)
deliver_data(data)
sndpkt = make_pkt(ACK, chksum)
udt_send(sndpkt)
Wait for
1 from
below
rdt_rcv(rcvpkt) && notcorrupt(rcvpkt)
&& has_seq0(rcvpkt)
extract(rcvpkt,data)
deliver_data(data)
sndpkt = make_pkt(ACK, chksum)
udt_send(sndpkt)
rdt_rcv(rcvpkt) && corrupt(rcvpkt)
sndpkt = make_pkt(NAK, chksum)
udt_send(sndpkt)
rdt_rcv(rcvpkt) && corrupt(rcvpkt)
sndpkt = make_pkt(NAK, chksum)
udt_send(sndpkt)
rdt2.1: receiver, handles garbled ACK/NAKs
Bit error in pkt
Transport Layer 3-47
rdt2.1: sender, handles garbled ACK/NAKs
Wait for
call 0 from
above
rdt_send(data)
sndpkt = make_pkt(0, data, checksum)
udt_send(sndpkt)
Wait for
ACK or
NAK 0 udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
isNAK(rcvpkt)
rdt_send(data)
sndpkt = make_pkt(1, data, checksum)
udt_send(sndpkt)
rdt_rcv(rcvpkt)
&& notcorrupt(rcvpkt)
&& isACK(rcvpkt)
udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
isNAK(rcvpkt)
rdt_rcv(rcvpkt)
&& notcorrupt(rcvpkt)
&& isACK(rcvpkt)
Wait for
call 1 from
above
Wait for
ACK or
NAK 1
LL
Re-send which pkt?
Transport Layer 3-48
Wait for
0 from
below
rdt_rcv(rcvpkt) && notcorrupt(rcvpkt)
&& has_seq1(rcvpkt)
extract(rcvpkt,data)
deliver_data(data)
sndpkt = make_pkt(ACK, chksum)
udt_send(sndpkt)
Wait for
1 from
below
rdt_rcv(rcvpkt) && notcorrupt(rcvpkt)
&& has_seq0(rcvpkt)
extract(rcvpkt,data)
deliver_data(data)
sndpkt = make_pkt(ACK, chksum)
udt_send(sndpkt)
rdt_rcv(rcvpkt) && corrupt(rcvpkt)
sndpkt = make_pkt(NAK, chksum)
udt_send(sndpkt)
rdt_rcv(rcvpkt) && corrupt(rcvpkt)
sndpkt = make_pkt(NAK, chksum)
udt_send(sndpkt)
rdt2.1: receiver, handles garbled ACK/NAKs
If bit error in pkt again
And again!!!
Re-sending loop!!!
If successfull
Unfortunately!ACK is corrupt
Transport Layer 3-49
rdt2.1: sender, handles garbled ACK/NAKs
Wait for
call 0 from
above
rdt_send(data)
sndpkt = make_pkt(0, data, checksum)
udt_send(sndpkt)
Wait for
ACK or
NAK 0 udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
( corrupt(rcvpkt) ||
isNAK(rcvpkt) )
rdt_send(data)
sndpkt = make_pkt(1, data, checksum)
udt_send(sndpkt)
rdt_rcv(rcvpkt)
&& notcorrupt(rcvpkt)
&& isACK(rcvpkt)
udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
( corrupt(rcvpkt) ||
isNAK(rcvpkt) )
rdt_rcv(rcvpkt)
&& notcorrupt(rcvpkt)
&& isACK(rcvpkt)
Wait for
call 1 from
above
Wait for
ACK or
NAK 1
LL
If the ACK/NAK is corrupted
Transport Layer 3-50
Wait for
0 from
below
rdt_rcv(rcvpkt) && notcorrupt(rcvpkt)
&& has_seq1(rcvpkt)
extract(rcvpkt,data)
deliver_data(data)
sndpkt = make_pkt(ACK, chksum)
udt_send(sndpkt)
Wait for
1 from
below
rdt_rcv(rcvpkt) && notcorrupt(rcvpkt)
&& has_seq0(rcvpkt)
extract(rcvpkt,data)
deliver_data(data)
sndpkt = make_pkt(ACK, chksum)
udt_send(sndpkt)
rdt_rcv(rcvpkt) && corrupt(rcvpkt)
sndpkt = make_pkt(NAK, chksum)
udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
not corrupt(rcvpkt) &&
has_seq0(rcvpkt)
sndpkt = make_pkt(ACK, chksum)
udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
not corrupt(rcvpkt) &&
has_seq1(rcvpkt)
sndpkt = make_pkt(ACK, chksum)
udt_send(sndpkt)
rdt_rcv(rcvpkt) && corrupt(rcvpkt)
sndpkt = make_pkt(NAK, chksum)
udt_send(sndpkt)
rdt2.1: receiver, handles garbled ACK/NAKs
If bit error in ACK again
And again!!!
Re-sending loop!!!
Transport Layer 3-51
rdt2.1: discussion
sender:
seq # added to pkt
two seq. #’s (0,1) will suffice. Why?
must check if received ACK/NAK corrupted
twice as many states state must “remember” whether “expected” pkt should have seq # of 0 or 1
receiver:
must check if received packet is duplicate state indicates whether
0 or 1 is expected pkt seq #
note: receiver can notknow if its last ACK/NAK received OK at sender
Transport Layer 3-52
rdt2.2: a NAK-free protocol
same functionality as rdt2.1, using ACKs only
instead of NAK, receiver sends ACK for last pkt received OK receiver must explicitly include seq # of pkt being ACKed
duplicate ACK at sender results in same action as NAK: retransmit current pkt
Transport Layer 3-53
rdt2.2: sender, receiver fragments
Wait for
call 0 from
above
sndpkt = make_pkt(0, data, checksum)
udt_send(sndpkt)
rdt_send(data)
udt_send(sndpkt)
rdt_rcv(rcvpkt) &&
( corrupt(rcvpkt) ||
isACK(rcvpkt,1) )
rdt_rcv(rcvpkt)
&& notcorrupt(rcvpkt)
&& isACK(rcvpkt,0)
Wait for
ACK
0
sender FSMfragment
rdt_rcv(rcvpkt) && notcorrupt(rcvpkt)
&& has_seq1(rcvpkt)
extract(rcvpkt,data)
deliver_data(data)
sndpkt = make_pkt(ACK1, chksum)
udt_send(sndpkt)
Wait for
0 from
below
rdt_rcv(rcvpkt) &&
(corrupt(rcvpkt) ||
has_seq1(rcvpkt))
udt_send(sndpkt)
receiver FSMfragment
L
Transport Layer 3-54
rdt3.0: channels with errors and loss
new assumption:underlying channel can also lose packets (data, ACKs)
checksum, seq. #, ACKs, retransmissions will be of help … but not enough
approach: sender waits “reasonable” amount of time for ACK
retransmits if no ACK received in this time
if pkt (or ACK) just delayed (not lost):
retransmission will be duplicate, but seq. #’s already handles this
receiver must specify seq # of pkt being ACKed
requires countdown timer
Transport Layer 3-55
rdt3.0 sender
sndpkt = make_pkt(0, data, checksum)
udt_send(sndpkt)
start_timer
rdt_send(data)
Wait
for
ACK0
rdt_rcv(rcvpkt) &&
( corrupt(rcvpkt) ||
isACK(rcvpkt,1) )
Wait for
call 1 from
above
sndpkt = make_pkt(1, data, checksum)
udt_send(sndpkt)
start_timer
rdt_send(data)
rdt_rcv(rcvpkt)
&& notcorrupt(rcvpkt)
&& isACK(rcvpkt,0)
rdt_rcv(rcvpkt) &&
( corrupt(rcvpkt) ||
isACK(rcvpkt,0) )
rdt_rcv(rcvpkt)
&& notcorrupt(rcvpkt)
&& isACK(rcvpkt,1)
stop_timer
stop_timer
udt_send(sndpkt)
start_timer
timeout
udt_send(sndpkt)
start_timer
timeout
rdt_rcv(rcvpkt)
Wait for
call 0from
above
Wait
for
ACK1
L
rdt_rcv(rcvpkt)
L
L
L
Transport Layer 3-56
sender receiver
rcv pkt1
rcv pkt0
send ack0
send ack1
send ack0
rcv ack0
send pkt0
send pkt1
rcv ack1
send pkt0
rcv pkt0pkt0
pkt0
pkt1
ack1
ack0
ack0
(a) no loss
sender receiver
rcv pkt1
rcv pkt0
send ack0
send ack1
send ack0
rcv ack0
send pkt0
send pkt1
rcv ack1
send pkt0
rcv pkt0pkt0
pkt0
ack1
ack0
ack0
(b) packet loss
pkt1X
loss
pkt1timeout
resend pkt1
rdt3.0 in action
Transport Layer 3-57
rdt3.0 in action
rcv pkt1send ack1
(detect duplicate)
pkt1
sender receiver
rcv pkt1
rcv pkt0
send ack0
send ack1
send ack0
rcv ack0
send pkt0
send pkt1
rcv ack1
send pkt0
rcv pkt0pkt0
pkt0
ack1
ack0
ack0
(c) ACK loss
ack1X
loss
pkt1timeout
resend pkt1
rcv pkt1send ack1
(detect duplicate)
pkt1
sender receiver
rcv pkt1
send ack0rcv ack0
send pkt1
send pkt0
rcv pkt0pkt0
ack0
(d) premature timeout/ delayed ACK
pkt1timeout
resend pkt1
ack1
send ack1
send pkt0rcv ack1
pkt0
ack1
ack0
send pkt0rcv ack1 pkt0
rcv pkt0send ack0ack0
rcv pkt0
send ack0(detect duplicate)
If packet delayed multiple timeouts
Transport Layer 3-58
rdt3.0: stop-and-wait operation
first packet bit transmitted, t = 0
sender receiver
RTT
last packet bit transmitted, t = L / R
first packet bit arrives
last packet bit arrives, send ACK
ACK arrives, send next
packet, t = RTT + L / R
Transport Layer 3-59
Performance of rdt3.0
rdt3.0 is correct, but performance stinks
e.g.: 1 Gbps link, 15 ms prop. delay, 8000 bit packet:
U sender: utilization – fraction of time sender busy sending
U sender =
.008
30.008 = 0.00027
L / R
RTT + L / R =
if RTT=30 msec, 1KB pkt every 30 msec: 33kB/sec thruput over 1 Gbps link
network protocol limits use of physical resources!
LR
8000 bits
109 bits/sec= = 8 microsecsDtrans =
Transport Layer 3-60
Pipelined protocols
pipelining: sender allows multiple, “in-flight”, yet-to-be-acknowledged pkts range of sequence numbers must be increased
buffering at sender and/or receiver
two generic forms of pipelined protocols: go-Back-N, selective repeat
Transport Layer 3-61
Pipelining: increased utilization
first packet bit transmitted, t = 0
sender receiver
RTT
last bit transmitted, t = L / R
first packet bit arrives
last packet bit arrives, send ACK
ACK arrives, send next
packet, t = RTT + L / R
last bit of 2nd packet arrives, send ACKlast bit of 3rd packet arrives, send ACK
3-packet pipelining increases
utilization by a factor of 3!
U sender =
.0024
30.008 = 0.00081
3L / R
RTT + L / R =
Transport Layer 3-62
Pipelined protocols: overview
Go-back-N: sender can have up to
N unacked packets in pipeline
receiver only sends cumulative ack doesn’t ack packet if
there’s a gap
sender has timer for oldest unacked packet when timer expires,
retransmit all unacked packets
Selective Repeat: sender can have up to N
unack’ed packets in pipeline
receiver sends individual ack for each packet
sender maintains timer for each unacked packet when timer expires,
retransmit only that unacked packet
Transport Layer 3-63
Go-Back-N: sender
k-bit seq # in pkt header
“window” of up to N, consecutive unack’ed pkts allowed
ACK(n): ACKs all pkts up to, including seq # n - “cumulative ACK” may receive duplicate ACKs (see receiver)
timer for oldest in-flight pkt
timeout(n): retransmit packet n and all higher seq # pkts in window
Transport Layer 3-64
GBN in action
send pkt0send pkt1send pkt2send pkt3
(wait)
sender receiver
receive pkt0, send ack0receive pkt1, send ack1
receive pkt3, discard, (re)send ack1rcv ack0, send pkt4
rcv ack1, send pkt5
pkt 2 timeoutsend pkt2send pkt3send pkt4send pkt5
Xloss
receive pkt4, discard, (re)send ack1
receive pkt5, discard, (re)send ack1
rcv pkt2, deliver, send ack2rcv pkt3, deliver, send ack3rcv pkt4, deliver, send ack4rcv pkt5, deliver, send ack5
ignore duplicate ACK
0 1 2 3 4 5 6 7 8
sender window (N=4)
0 1 2 3 4 5 6 7 8
0 1 2 3 4 5 6 7 8
0 1 2 3 4 5 6 7 8
0 1 2 3 4 5 6 7 8
0 1 2 3 4 5 6 7 8
0 1 2 3 4 5 6 7 8
0 1 2 3 4 5 6 7 8
0 1 2 3 4 5 6 7 8
0 1 2 3 4 5 6 7 8
Transport Layer 3-65
GBN: sender extended FSM
Waitstart_timer
udt_send(sndpkt[base])
udt_send(sndpkt[base+1])
…
udt_send(sndpkt[nextseqnum-1])
timeout
rdt_send(data)
if (nextseqnum < base+N) {
sndpkt[nextseqnum] = make_pkt(nextseqnum,data,chksum)
udt_send(sndpkt[nextseqnum])
if (base == nextseqnum)
start_timer
nextseqnum++
}
else
refuse_data(data)
base = getacknum(rcvpkt)+1
If (base == nextseqnum)
stop_timer
else
start_timer
rdt_rcv(rcvpkt) &&
notcorrupt(rcvpkt)
base=1
nextseqnum=1
rdt_rcv(rcvpkt)
&& corrupt(rcvpkt)
L
What if a duplicated ACK?
When to stop/start timer
Draw the sliding window!
What if a out-of-order ACK?
Transport Layer 3-66
ACK-only: always send ACK for correctly-received pkt with highest in-order seq # may generate duplicate ACKs
need only remember expectedseqnum
out-of-order pkt: discard (don’t buffer): no receiver buffering! re-ACK pkt with highest in-order seq #
Wait
udt_send(sndpkt)
default
rdt_rcv(rcvpkt)
&& notcurrupt(rcvpkt)
&& hasseqnum(rcvpkt,expectedseqnum)
extract(rcvpkt,data)
deliver_data(data)
sndpkt = make_pkt(expectedseqnum,ACK,chksum)
udt_send(sndpkt)
expectedseqnum++
expectedseqnum=1
sndpkt =
make_pkt(expectedseqnum,ACK,chksum)
L
GBN: receiver extended FSM
Transport Layer 3-67
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 limits seq #s of sent, unACKed pkts
Transport Layer 3-68
Selective repeat: sender, receiver windows
Transport Layer 3-69
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 #
sender
How is the window sliding?
Transport Layer 3-70
Selective repeat
pkt 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
Transport Layer 3-71
Selective repeat in action
send pkt0send pkt1send pkt2send pkt3
(wait)
sender receiver
receive pkt0, send ack0receive pkt1, send ack1
receive pkt3, buffer, send ack3rcv ack0, send pkt4
rcv ack1, send pkt5
pkt 2 timeoutsend pkt2
Xloss
receive pkt4, buffer, send ack4
receive pkt5, buffer, send ack5
rcv pkt2; deliver pkt2,pkt3, pkt4, pkt5; send ack2
record ack3 arrived
0 1 2 3 4 5 6 7 8
sender window (N=4)
0 1 2 3 4 5 6 7 8
0 1 2 3 4 5 6 7 8
0 1 2 3 4 5 6 7 8
0 1 2 3 4 5 6 7 8
0 1 2 3 4 5 6 7 8
0 1 2 3 4 5 6 7 8
0 1 2 3 4 5 6 7 8
0 1 2 3 4 5 6 7 8
0 1 2 3 4 5 6 7 8
record ack4 arrived
record ack5 arrived
Q: what happens when ack2 arrives?
Transport Layer 3-72
Selective repeat:dilemma
example: seq #’s: 0, 1, 2, 3 window size=3
receiver window(after receipt)
sender window(after receipt)
0 1 2 3 0 1 2
0 1 2 3 0 1 2
0 1 2 3 0 1 2
pkt0
pkt1
pkt2
0 1 2 3 0 1 2 pkt0
timeoutretransmit pkt0
0 1 2 3 0 1 2
0 1 2 3 0 1 2
0 1 2 3 0 1 2X
X
X
will accept packetwith seq number 0
(b) oops!
0 1 2 3 0 1 2
0 1 2 3 0 1 2
0 1 2 3 0 1 2
pkt0
pkt1
pkt2
0 1 2 3 0 1 2
pkt0
0 1 2 3 0 1 2
0 1 2 3 0 1 2
0 1 2 3 0 1 2
X
will accept packetwith seq number 0
0 1 2 3 0 1 2 pkt3
(a) no problem
receiver can’t see sender side.receiver behavior identical in both cases!
something’s (very) wrong!
receiver sees no difference in two scenarios!
duplicate data accepted as new in (b)
Q: what relationship between seq # size and window size to avoid problem in (b)?
Transport Layer
Animation
Transport Layer 3-74
Chapter 3 outline
3.1 transport-layer services
3.2 multiplexing and demultiplexing
3.3 connectionless transport: UDP
3.4 principles of reliable data transfer
3.5 connection-oriented transport: TCP
segment structure
reliable data transfer
flow control
connection management
3.6 principles of congestion control
3.7 TCP congestion control
Transport Layer 3-75
TCP: Overview RFCs: 793,1122,1323, 2018, 2581
full duplex data: bi-directional data flow
in same connection
MSS: maximum segment size
connection-oriented: handshaking (exchange
of control msgs) inits sender, receiver state before data exchange
flow controlled: sender will not
overwhelm receiver
point-to-point: one sender, one receiver
reliable, in-order byte steam: no “message
boundaries”
pipelined: TCP congestion and
flow control set window size
Transport Layer 3-76
TCP segment structure
source port # dest port #
32 bits
application
data
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUhead
len
not
used
options (variable length)
URG: urgent data
(generally not used)
ACK: ACK #
valid
PSH: push data now
(generally not used)
RST, SYN, FIN:
connection estab
(setup, teardown
commands)
# bytes
rcvr willing
to accept
counting
by bytes
of data
(not segments!)
Internet
checksum
(as in UDP)
Transport Layer 3-77
TCP seq. numbers, ACKs
sequence numbers:
byte stream “number” of first byte in segment’s data
acknowledgements:
seq # of next byte expected from other side
cumulative ACK
Q: how receiver handles out-of-order segments
A: TCP spec doesn’t say, -up to implementor source port # dest port #
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent, not-yet ACKed(“in-flight”)
usablebut not yet sent
not usable
window sizeN
sender sequence number space
source port # dest port #
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-78
TCP seq. numbers, ACKs
Usertypes‘C’
host ACKsreceipt
of echoed‘C’
host ACKsreceipt of‘C’, echoesback ‘C’
simple telnet scenario
Host BHost A
Seq=42, ACK=79, data = ‘C’
Seq=79, ACK=43, data = ‘C’
Seq=43, ACK=80
Transport Layer 3-79
TCP round trip time, timeout
Q: how to set TCP timeout value?
longer than RTT
but RTT varies
too short: premature timeout, unnecessary retransmissions
too long: slow reaction to segment loss
Q: how to estimate RTT? SampleRTT: measured
time from segment transmission until ACK receipt
ignore retransmissions
SampleRTT will vary, want estimated RTT “smoother” average several recent
measurements, not just current SampleRTT
Wasting time
Wasting bandwidth
Transport Layer 3-80
RTT: gaia.cs.umass.edu to fantasia.eurecom.fr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RT
T (
mil
lise
con
ds)
SampleRTT Estimated RTT
EstimatedRTT = (1- α)*EstimatedRTT + α *SampleRTT
exponential weighted moving average influence of past sample decreases exponentially fast typical value: α = 0.125
TCP round trip time, timeout
RTT (milliseconds)
RTT: gaia.cs.umass.edu to fantasia.eurecom.fr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-81
timeout interval: EstimatedRTT plus “safety margin” large variation in EstimatedRTT -> larger safety margin
estimate SampleRTT deviation from EstimatedRTT:
DevRTT = (1- β)*DevRTT +
β *|SampleRTT-EstimatedRTT|
TCP round trip time, timeout
(typically, β = 0.25)
TimeoutInterval = EstimatedRTT + 4*DevRTT
estimated RTT “safety margin”
Transport Layer 3-82
Chapter 3 outline
3.1 transport-layer services
3.2 multiplexing and demultiplexing
3.3 connectionless transport: UDP
3.4 principles of reliable data transfer
3.5 connection-oriented transport: TCP
segment structure
reliable data transfer
flow control
connection management
3.6 principles of congestion control
3.7 TCP congestion control
Transport Layer 3-83
TCP reliable data transfer
TCP creates rdt service on top of IP’s unreliable service pipelined segments
cumulative acks
single retransmission timer
retransmissions triggered by: timeout events
duplicate acks
let’s initially consider simplified TCP sender: ignore duplicate acks
ignore flow control, congestion control
Transport Layer 3-84
TCP sender events:
data rcvd from app:
create segment with seq #
seq # is byte-stream number of first data byte in segment
start timer if not already running think of timer as for
oldest unacked segment
expiration interval: TimeOutInterval
timeout:
retransmit segment that caused timeout
restart timer
ack rcvd:
if ack acknowledges previously unacked segments update what is known
to be ACKed
start timer if there are still unacked segments
Transport Layer 3-85
TCP sender (simplified)
wait
for
event
NextSeqNum = InitialSeqNum
SendBase = InitialSeqNum
L
create segment, seq. #: NextSeqNum
pass segment to IP (i.e., “send”)
NextSeqNum = NextSeqNum + length(data)
if (timer currently not running)
start timer
data received from application above
retransmit not-yet-acked segment with smallest seq. #
start timer
timeout
if (y > SendBase) {
SendBase = y
/* SendBase–1: last cumulatively ACKed byte */
if (there are currently not-yet-acked segments)
start timer
else stop timer
}
ACK received, with ACK field value y
Transport Layer 3-86
TCP: retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92, 8 bytes of data
ACK=100
Seq=92, 8 bytes of data
Xtimeout
ACK=100
premature timeout
Host BHost A
Seq=92, 8 bytes of data
ACK=100
Seq=92, 8bytes of data
timeout
ACK=120
Seq=100, 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-873-87
TCP: retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92, 8 bytes of data
ACK=100
Seq=120, 15 bytes of data
timeout
Seq=100, 20 bytes of data
ACK=120
Do we need to retransmit the first pkt?
The sequent timeout is doubled
until receiving an ACK or rdt_sent?
Transport Layer 3-883-88
TCP ACK generation [RFC 1122, RFC 2581]
event at receiver
arrival of in-order segment with
expected seq #. All data up to
expected seq # already ACKed
arrival of in-order segment with
expected seq #. One other
segment has ACK pending
arrival of out-of-order segment
higher-than-expect seq. # .
Gap detected
arrival of segment that
partially or completely fills gap
TCP receiver action
delayed ACK. Wait up to 500ms
for next segment. If no next segment,
send ACK
immediately send single cumulative
ACK, ACKing both in-order segments
immediately send duplicate ACK,
indicating seq. # of next expected byte
immediate send ACK, provided that
segment starts at lower end of gap
Transport Layer 3-89
Problem!
Transport Layer 3-90
X
Host BHost A
Seq=92, 8 bytes of data
ACK=100
timeout ACK=100
ACK=100
ACK=100
Problems
Seq=100, 20 bytes of data
Seq=100, 20 bytes of data
Transport Layer 3-91
TCP fast retransmit
time-out period often relatively long: long delay before
resending lost packet
detect lost segments via duplicate ACKs. sender often sends
many segments back-to-back
if segment is lost, there will likely be many duplicate ACKs.
if sender receives 3 ACKs for same data
(“triple duplicate ACKs”),resend unacked segment with smallest seq # likely that unacked
segment lost, so don’t wait for timeout
TCP fast retransmit
Transport Layer 3-92
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92, 8 bytes of data
ACK=100
timeout ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100, 20 bytes of data
Seq=100, 20 bytes of data
Transport Layer 3-93
Chapter 3 outline
3.1 transport-layer services
3.2 multiplexing and demultiplexing
3.3 connectionless transport: UDP
3.4 principles of reliable data transfer
3.5 connection-oriented transport: TCP
segment structure
reliable data transfer
flow control
connection management
3.6 principles of congestion control
3.7 TCP congestion control
Transport Layer 3-94
TCP flow control
Speed-Matching Service Match Send Rate to Receiving App’s Drain Rate
Transport Layer 3-95
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers ….
… slower than TCP receiver is delivering(sender is sending)
from sender
receiver controls sender, so
sender won’t overflow receiver’s buffer by transmitting too much, too fast
flow control
Transport Layer 3-96
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver “advertises” free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size set via
socket options (typical default is 4096 bytes)
many operating systems autoadjust RcvBuffer
sender limits amount of unacked (“in-flight”) data to receiver’s rwnd value
guarantees receive buffer will not overflow
receiver-side buffering
What if rwnd = 0 !
Transport Layer 3-97
Chapter 3 outline
3.1 transport-layer services
3.2 multiplexing and demultiplexing
3.3 connectionless transport: UDP
3.4 principles of reliable data transfer
3.5 connection-oriented transport: TCP
segment structure
reliable data transfer
flow control
connection management
3.6 principles of congestion control
3.7 TCP congestion control
Transport Layer 3-98
Connection Management
before exchanging data, sender/receiver “handshake”: agree to establish connection (each knowing the other willing
to establish connection)
agree on connection parameters
connection state: ESTABconnection variables:
seq # client-to-serverserver-to-client
rcvBuffer size
at server,client
application
network
connection state: ESTABconnection Variables:
seq # client-to-serverserver-to-client
rcvBuffer size
at server,client
application
network
Socket clientSocket =
newSocket("hostname","port
number");
Socket connectionSocket =
welcomeSocket.accept();
Transport Layer 3-99
Q: will 2-way handshake always work in network?
variable delays
retransmitted messages (e.g. req_conn(x)) due to message loss
message reordering
can’t “see” other side
2-way handshake:
Let’s talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-100
Agreeing to establish a connection
2-way handshake failure scenarios:
retransmitreq_conn(x)
ESTAB
req_conn(x)
half open connection!(no client!)
client terminates
serverforgets x
connection x completes
retransmitreq_conn(x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminates
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-101
TCP 3-way handshake
SYNbit=1, Seq=x
choose init seq num, xsend TCP SYN msg
ESTAB
SYNbit=1, Seq=yACKbit=1; ACKnum=x+1
choose init seq num, ysend TCP SYNACKmsg, acking SYN
ACKbit=1, ACKnum=y+1
received SYNACK(x) indicates server is live;send ACK for SYNACK;
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-102
TCP 3-way handshake: FSM
closed
L
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket =
newSocket("hostname","port
number");
SYN(seq=x)
Socket connectionSocket =
welcomeSocket.accept();
SYN(x)
SYNACK(seq=y,ACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=y,ACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
L
Server Client
Transport Layer 3-103
SYN Flood Attack
Deluge of SYNs
Deplete Server Resources Half-Open Connections
SYN Cookies
Don’t Allocate Resources until ACK Received
Transport Layer 3-104
TCP: closing a connection
client, server each close their side of connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN, ACK can be combined with own FIN
simultaneous FIN exchanges can be handled
Transport Layer 3-105
FIN_WAIT_2
CLOSE_WAIT
FINbit=1, seq=y
ACKbit=1; ACKnum=y+1
ACKbit=1; ACKnum=x+1
wait for serverclose
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2*max
segment lifetime
CLOSED
TCP: closing a connection
FIN_WAIT_1 FINbit=1, seq=xcan no longersend but canreceive data
clientSocket.close()
client state server state
ESTABESTAB
Transport Layer 3-106
Chapter 3 outline
3.1 transport-layer services
3.2 multiplexing and demultiplexing
3.3 connectionless transport: UDP
3.4 principles of reliable data transfer
3.5 connection-oriented transport: TCP
segment structure
reliable data transfer
flow control
connection management
GBN or SR
3.6 principles of congestion control
3.7 TCP congestion control
Transport Layer 3-107
What is congestion control?
Basic approaches to congestion control
In the network
From the end host
Principles of congestion control
Transport Layer 3-108
Congestion What is it?
What isn’t it?
Problems Lost Packets
Long Delays
Principles of congestion control
Transport Layer 3-109
congestion: informally: “too many sources sending too much
data too fast for network to handle” different from flow control!
manifestations:
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem!
Principles of congestion control
Transport Layer 3-110
Two packets colliding at a router, heading the same port
Flows using up all link capacity (TCP communication flow)
Too many users using a link during a peak hour
Time Scales of Congestion
queue
Trans. Rate
Multi. RTT
What causes congestion?
t
Cum
ula
tive b
its
12Mb/s
A1(t)
A1(t)+A2(t)
Q(t)
Q(t) What if the buffer is finite?
Another example?
Are there congestions?
Notice the sender to avoid downstream pkt drop!
How to split the trans. Rate?
6 Mb/s
6 Mb/s
A:3 Mb/s
C:6 Mb/s
B:3 Mb/s
A:4 Mb/s
C:4 Mb/s
B:4 Mb/s
Another example?
Are there congestions?
Notice the sender to avoid downstream pkt drop!
How to split the trans. Rate?
A:4 Mb/s
C:4 Mb/s
B:4 Mb/s
D:? Mb/s
Congestion is unavoidable
Arguably a little congestion is good! : We use packet switching because it makes efficient use of
the links. Therefore, buffers in the routers are frequently occupied.
If buffers are always empty, delay is low, but our usage of the network is low
If buffers are always occupied, delay is high, but we are using the network more efficiently
Observations
Congestion is inevitable, and arguably desirable.
Congestion happens at different time scales - from two individual packets colliding, to some flows sending too quickly, to flash crowds appearing in the network.
If packets are dropped, then retransmissions can make congestion even worse.
When packets are dropped, they waste resources upstream before they were dropped.
We need a definition of fairness, to decide how we want flows to share a bottleneck link.
Fairness and throughput
How to allocate the trans. rate if everyone wants to send at the maximum rate
2 1A
B C
0.25
1.75 0.75
Total throughput = 2.75
0.5
1.5 0.5
Total throughput = 2.5
In first allocation, A is penalized for achieving a higher throughput. Therefore, it’s a tradeoff between throughput and fairness
Max-min Fairness
Definition:
An allocation is max-min fair if you can not increase the rate of one flow without decreasing the rate of another flow with a lower rate
2 1A
B C
0.5
1.5 0.5
Total throughput = 2.5
Curtailing for equal rate in bottleneck
Max-min Fairness: Single link
Definition is intuitive and simple on a single link.
1
A
B
C
0.5
1
0.2
0.2
0.4
0.4 Max-min fair
Goals for congestion control
High throughput: keep links busy and flows fast
Max-min fairness
Respond quickly to changing network conditions
Distributed control
Where to put congestion control?
Example: FQ at every router
R
R/2
R/2
Not responsible, cannot tell the sender how the rate should be set!!
Transport Layer 3-121
Causes/costs of congestion: scenario 1
two senders, two receivers
one router, infinite buffers
output link capacity: R
no retransmission
maximum per-connection throughput: R/2
unlimited shared
output link buffers
Host A
original data: lin
Host B
throughput: lout
R/2
R/2
lout
lin R/2dela
ylin
large delays as arrival rate, lin, approaches capacity
Queueing Delay
Transport Layer 3-122
Causes/costs of congestion: scenario 1
Large queueing DELAY are experienced as the packet arrival rate nears the link
capacity
Transport Layer 3-123
one router, finite buffers
sender retransmission of timed-out packet application-layer input = application-layer output: lin = lout
transport-layer input includes retransmissions : lin lin
finite shared output
link buffers
Host A
lin : original data
Host B
loutl'in: original data, plusretransmitted data
‘
Causes/costs of congestion: scenario 2
Transport Layer 3-124
idealization: sender has perfect knowledge
sender sends only when router buffers available
finite shared output
link buffers
lin : original dataloutl'in: original data, plus
retransmitted data
copy
free buffer space!
R/2
R/2
lout
lin
Causes/costs of congestion: scenario 2
Host B
A
Transport Layer 3-125
lin : original dataloutl'in: original data, plus
retransmitted data
copy
no buffer space!
Idealization: known losspackets can be lost, dropped at router due to full buffers
sender only resends if packet known to be lost
Causes/costs of congestion: scenario 2
A
Host B
Transport Layer 3-126
lin : original dataloutl'in: original data, plus
retransmitted data
free buffer space!
Causes/costs of congestion: scenario 2
Idealization: known losspackets can be lost, dropped at router due to full buffers
sender only resends if packet known to be lost
R/2
R/2lin
lout
when sending at R/2,
some packets are
retransmissions but
asymptotic goodput
is still R/2 (why?)
A
Host B
Transport Layer 3-127
A
linloutl'in
copy
free buffer space!
timeout
R/2
R/2lin
lout
when sending at R/2,
some packets are
retransmissions
including duplicated
that are delivered!
Host B
Realistic: duplicates packets can be lost, dropped
at router due to full buffers
sender times out prematurely, sending two copies, both of which are delivered
Causes/costs of congestion: scenario 2
Transport Layer 3-128
R/2
lout
when sending at R/2,
some packets are
retransmissions
including duplicated
that are delivered!
“costs” of congestion: more work (retrans) for given “goodput” unneeded retransmissions: link carries multiple copies of pkt
decreasing goodput
R/2lin
Causes/costs of congestion: scenario 2
Realistic: duplicates packets can be lost, dropped
at router due to full buffers
sender times out prematurely, sending two copies, both of which are delivered
Transport Layer 3-129
four senders
multihop paths
timeout/retransmit
Q: what happens as lin and lin’
increase ?
finite shared output
link buffers
Host A lout
Causes/costs of congestion: scenario 3
Host B
Host C
Host D
lin : original data
l'in: original data, plusretransmitted data
A: as red lin’ increases, all arriving
blue pkts at upper queue are dropped, blue throughput g 0
Transport Layer 3-130
another “cost” of congestion:
when packet dropped, any “upstream transmission capacity used for that packet was wasted!
Causes/costs of congestion: scenario 3
C/2
C/2
lout
lin’
A B
D C
Transport Layer 3-131
Approaches towards congestion control
two broad approaches towards congestion control:
end-end congestion control:
no explicit feedback from network
congestion inferred from end-system observed loss, delay
approach taken by TCP
network-assisted congestion control:
routers provide feedback to end systems
single bit indicating congestion (SNA, DECbit, TCP/IP ECN, ATM)
explicit rate for sender to send at
Network-based
Q:Who is responsible ?
A: Router
Q:How to indicate a congestion?
A: Pkt drop, queue occupancy, threshold, link capacity etc.
Q:How to get back to sender?
TCP: ECN, DECbit(data cost)
Congestion!!!
DECbit
Transport Layer 3-133
Case study: ATM ABR congestion control
ABR: available bit rate: “elastic service”
if sender’s path “underloaded”: sender should use
available bandwidth
if sender’s path congested:
sender throttled to minimum guaranteed rate
RM (resource management) cells:
sent by sender, interspersed with data cells
bits in RM cell set by switches (“network-assisted”) NI bit: no increase in rate
(mild congestion)
CI bit: congestion indication
RM cells returned to sender by receiver, with bits intact
Transport Layer 3-134
Case study: ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell
senders’ send rate thus max supportable rate on path
EFCI bit in data cells: set to 1 in congested switch
if data cell preceding RM cell has EFCI set, receiver sets CI bit in returned RM cell
RM cell data cell
End-host based
Q: Is it worth providing congestion control by the network?
Q:How can the host observe the network behavior (congestion)?
A: drop, delay …
For TCP, it has to do this because IP provides no support
Transport Layer 3-136
Overload sliding window mechanism
Exploits TCP’s sliding window used for flow control.
Tries to figure out how many packets it can safely have outstanding in the network at a time.
sent ACKed
sent, not-yet ACKed(“in-flight”)
usablebut not yet sent
not usable
sender sequence number space
Transport Layer 3-137
AIMD: additive increase multiplicative decrease
Transport Layer 3-138
Animation for AIMDhttp://guido.appenzeller.net/anims/
Transport Layer 3-139
Single flow dynamicslockstep
Transport Layer 3-140
Sending rate for single flow
Constant!!!
How large should the buffer be?
Transport Layer 3-141
Sending rate for single flow
Transport Layer 3-142
Observations for single flow
Window expands/contracts according to AIMD
…to probe how many bytes the pipe can hold.
The sawtooth is the stable operating point.
The sending rate is constant.
…if we have sufficient buffers (RTT x C).
Transport Layer 3-143
Multiple flows
Transport Layer 3-144
One flow vs multiple flows
Why the saw tooth has different shape ??
Transport Layer 3-145
Simple geometric intuition
Transport Layer 3-146
Observations for multiple flow
Window expands/contracts according to AIMD.
…to probe how many bytes the pipe can hold.
Bottleneck will contain packets from many flows.
The sending rate varies with window size.
AIMD is very sensitive to loss rate.
AIMD penalizes flows with long RTTs.
Transport Layer 3-147
Chapter 3 outline
3.1 transport-layer services
3.2 multiplexing and demultiplexing
3.3 connectionless transport: UDP
3.4 principles of reliable data transfer
3.5 connection-oriented transport: TCP
segment structure
reliable data transfer
flow control
connection management
3.6 principles of congestion control
3.7 TCP congestion control
Transport Layer 3-148
TCP Congestion Control
End-to-End Reacts to events observable at the end host (e.g.
packet loss).
Limit send rate when network is congested
Questions How to perceive congestion?
How to limit send rate?
How to change send rate?
Transport Layer 3-149
How to Perceive Congestion?
Implicit End-to-End Feedback
ACK Received ?
ACK Not Received ?
Transport Layer 3-150
TCP Congestion Control: details
sender limits transmission:
cwnd is dynamic, function of perceived network congestion
TCP sending rate:
roughly: send cwnd bytes, wait RTT for ACKS, then send more bytes
last byteACKed sent, not-
yet ACKed(“in-flight”)
last byte sent
cwnd
LastByteSent-
LastByteAcked< cwnd
sender sequence number space
rate ~~cwnd
RTTbytes/sec
How to change cwnd ?
Transport Layer 3-151
How to Limit Send Rate?
Limiting # of unACKed bytes “in pipeline” cwnd: differs from rwnd (how, why?)
Sender limited by min(__, __)
LastByteSent - LastByteAcked
Transport Layer 3-152
TCP Congestion Control
Goal Fast But Not Too Fast
How to find rate just below congestion level?
Transport Layer 3-153
TCP congestion control (AIMD)
approach: sender increases transmission rate (window size), probing for usable bandwidth, until loss occurs
additive increase: increase cwnd by 1 MSS every RTT until loss detected
multiplicative decrease: cut cwnd in half after loss cwnd:
TC
P s
end
er
co
ng
estion w
ind
ow
siz
e
AIMD saw tooth
behavior: probing
for bandwidth
additively increase window size ……. until loss occurs (then cut window in half)
time
What happens in detail?
Transport Layer 3-154
Practical TCP Congestion control
Flow control window is only about endpoint
Have TCP estimate a congestion window for the network
Sender window = min(flow window, congestion window)
Separate congestion control into two states Slow start: on connection startup or packet
timeout
Congestion avoidance: steady operation
TCP Tahoe, 1987-1988
Transport Layer 3-155
Success Event
If ACK Received _______cwndincrease
Slowstart Increase Exponentiially
Connection start or after Timeout
Congestion Avoidance Increase linearly
Normal Operation
Increase by MSS2/congestion window for
each acknowledgment;
Increase by MSS each round trip time
Transport Layer 3-156
Loss Event
If Segment Lost _______cwndDecrease
Timeout Cut cwnd to 1
Duplicated ACKs Cut cwnd to half
Transport Layer 3-157
TCP Slow Start
when connection begins,increase rate exponentially until first loss event: initially cwnd = 1 MSS
double cwnd every RTT
done by incrementing cwnd for every ACK received
Feature: initial rate is slow but ramps up exponentially fast
Host A
RT
T
Host B
time
Transport Layer 3-158
TCP Slow Start
ssthresh Cwnd threshold maintained by TCP
On timeout Ssthresh = cwnd/2
When cwnd >= ssthresh Congestion Avoidance
Transport Layer 3-159
TCP Slow Start Summary
Slow start Window starts at
Maximum Segment Size (MSS)
Increase window by MSS for each acknowledged packet
Exponentially grow congestion window to sense network capacity
“Slow” compared to prior approach
Transport Layer 3-160
TCP: detecting, reacting to loss
loss indicated by timeout: cwnd set to 1 MSS;
window then grows exponentially (as in slow start) to threshold, then grows linearly
loss indicated by 3 duplicate ACKs: TCP RENO dup ACKs indicate network capable of delivering
some segments
cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-161
Q: when should the exponential increase switch to linear? Why?
A: when cwnd gets to 1/2 of its value before timeout.
A: Less aggressive
Implementation: variable ssthresh
on loss event, ssthresh is set to 1/2 of cwnd just before loss event
Cwnd += MSS/cwnd for each ACK received
TCP: switching from slow start to CA
Transport Layer 3-162
Summary: TCP Congestion Control
timeout
ssthresh = cwnd/2cwnd = 1 MSS
dupACKcount = 0
retransmit missing segment
L
cwnd > ssthresh
congestion
avoidance
cwnd = cwnd + MSS (MSS/cwnd)dupACKcount = 0
transmit new segment(s), as allowed
new ACK.
dupACKcount++
duplicate ACK
fast
recovery
cwnd = cwnd + MSStransmit new segment(s), as allowed
duplicate ACK
ssthresh= cwnd/2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeout
ssthresh = cwnd/2cwnd = 1 dupACKcount = 0
retransmit missing segment
ssthresh= cwnd/2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow
start
timeout
ssthresh = cwnd/2 cwnd = 1 MSS
dupACKcount = 0retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s), as allowed
new ACKdupACKcount++
duplicate ACK
L
cwnd = 1 MSSssthresh = 64 KBdupACKcount = 0
NewACK!
NewACK!
NewACK!
Transport Layer 3-163
TCP Congestion Control: Phases• cwnd < ssthresh ?
• cwnd > ssthresh ?
Transport Layer 3-164
TCP Congestion Control: Phases• cwnd < ssthresh ?
• cwnd > ssthresh ?
• Timeout ? • Triple Duplicated ACK ?
Transport Layer 3-165
TCP Congestion Control: Phases• Timeout ?
• Ssthresh = cwnd/2• Cwnd = 1MSS
• Triple Duplicated ACK ?
• Ssthresh = cwnd/2• Cwnd = ssthresh
Transport Layer 3-166
TCP throughput
avg. TCP thruput as function of window size, RTT? ignore slow start, assume always data to send
W: window size (measured in bytes) where loss occurs avg. window size (# in-flight bytes) is ¾ W
avg. thruput is 3/4W per RTT
W
W/2
avg TCP thruput = 34
WRTT
bytes/sec
Transport Layer 3-167
TCP Futures: TCP over “long, fat pipes”
example: 1500 byte segments, 100ms RTT, want 10 Gbps throughput
requires W = 83,333 in-flight segments
throughput in terms of segment loss probability, L [Mathis 1997]:
➜ to achieve 10 Gbps throughput, need a loss rate of L = 2·10-10 – a very small loss rate!
new versions of TCP for high-speed
TCP throughput = 1.22 . MSS
RTT L
Transport Layer 3-168
fairness goal: if K TCP sessions share same bottleneck link of bandwidth R, each should have average rate of R/K
TCP connection 1
bottleneck
router
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-169
Why is TCP fair?
two competing sessions: additive increase gives slope of 1, as throughout increases
multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughput
congestion avoidance: additive increase
loss: decrease window by factor of 2
congestion avoidance: additive increaseloss: decrease window by factor of 2
Transport Layer 3-170
Fairness (more)
Fairness and UDP
multimedia apps often do not use TCP do not want rate
throttled by congestion control
instead use UDP: send audio/video at
constant rate, tolerate packet loss
Fairness, parallel TCP connections
application can open multiple parallel connections between two hosts
web browsers do this
e.g., link of rate R with 9 existing connections: new app asks for 1 TCP, gets rate
R/10
new app asks for 11 TCPs, gets R/2
Transport Layer 3-171
Chapter 3: summary
principles behind transport layer services:
multiplexing, demultiplexing
reliable data transfer
flow control
congestion control
instantiation, implementation in the Internet UDP
TCP
next:
leaving the network “edge”(application, transport layers)
into the network “core”