Ch 3. Transport Layer Myungchul Kim

2 o Logical communication between application processes running on different hosts o A network-layer protocol provides logical communication between hosts. o A transport protocol can offer reliable data transfer service to an application even when the underlying network protocol is unreliable. o IP service model: best-effort delivery service o UDP: unregulated

3 – Fig 3.1

4 o Intermediate routers neither act on, nor recognize, any information that the transport layer may have added to the application messages. o Transport-layer multiplexing: extending host-to-host delivery to process-to-process delivery

5 o Host-to-host delivery vs process-to-process delivery o Multiplexing o Demultiplexing o Transport-layer multiplexing – Sockets have unique identifiers – Each segment have special fields, port number fields, that indicate the socket to which the segment is to be delivered. – Port number: 16-bit number: from 0 to 65535, well-known port numbers from 0 to e.g., HTTP (80), FTP(21) – Fig 3.3 Multiplexing and Demultiplexing

6 o A UDP socket is fully identified by a two-tuple consisting of a destination IP address and a destination port number. o Fig 3.4

7 o A TCP socket is identified by a 4-tuple: (source IP address, source port number, destination IP address, destination port number). o In contrast with UDP, two arriving TCP segments with different source IP addresses or source port numbers will be directed to two different sockets. o The HTTP server distinguishes the segments from the different clients by the source IP addresses and source port numbers. o Persistent HTTP vs Nonpersistent HTTP

8 o Fig 3.5

9 o Multiplexing and demultiplexing o Light error checking o The application is almost directly talking with IP. o Connectionless o No handshaking o Merits – No connection establishment -> no delay – No connection state – Small packet header overhead: TCP(20), UDP(8 bytes) – Finer application-level control over what data is sent, and when. o How about congestion due to UDP? – New mechanism to force all sources including UDP sources to perform adaptive congestion control. – Applications have built acknowledgements and retransmissions into the applications. UDP

10 o Fig 3.6

11 o UDP segment structure – Fig 3.7 o UDP checksum – For error detection – UDP at the sender side performs the 1’s complement of the sum of all the 16-bit words in the segment. – At the receiver, all four 16-bit words are added, including the checksum. – If no errors, If one of the bits is a zero, then errors. o No recovery from an error.

12 o The transport layer, the link layer and the application layer o Fig 3.8 Principles of Reliable Data Transfer

13 o Reliable data transfer over a Perfectly Reliable Channel: rdt1.0 – Separate FSM – Event/action – Initial state – Fig 3.9 – No difference between a unit of data and a packet – No feedback: reliable channel – No flow control

14 o Reliable data transfer over a Channel: rdt2.0 – Bits in a packet may be corrupted – All transmitted packets are received in the order in which they were sent. – Positive ack and negative ack – ARQ(Automatic Repeat reQuest) protocols – Capabilities handling the presence of bit errors:  Error detection  Receiver feedback: a 0 value indicate a NAK and a value of 1 could indicate an ACK  Retransmission

15 – Fig. 3.10

16 o Reliable data transfer over a Channel: rdt2.0 – When the receiver is in the wait-for-ACK-or-NAK state, it cannot get more data from the upper layer. – Stop-and-wait protocol – Drawbacks  The ACK or NAK packet could be corrupted. -> checksum  How the protocol should recover from errors in ACK or NAK packets. – A one-bit sequence number: allow the receiver to know whether the sender is sending the previously transmitted packet. – Since we are currently assuming a channel that does not lose packets, ACK and NAK packets do not themselves need to indicate the sequence number of the packet they are acknowledging. – When an out-of-order packet is received, the receiver sends a positive acknowledgement for the packet it has received. -> duplicate ACKs – When a corrupted packet is received, the receiver sends a negative acknowledgement.

17 – Fig. 3.11

18 – Fig. 3.12

19 o Reliable data transfer over a Lossy Channel with Bit Errors: rdt3.0 – How to detect packet loss and what to do when packet loss occurs – Use of checksuming, sequence numbers, ACK packets and retransmissions – Choose a time value such that packet loss is likely, although not guaranteed, to have happened. – Duplicate data packets. – A delayed ACK: with an acknowledgement field – Alternating-bit protocol

20 – Fig. 3.15

21 – Fig. 3.16

22 o RTT: 30msec between the coasts, t trans = L/R = 8000bits/packet / 10 9 bits/sec = 8 microsec o Utilization of the channel: U sender = L/R / RTT + L/R = o Effective throughput = 267kbps on 1Gbps o Pipelining : the sender is allowed to send multiple packets without waiting for acknowledgements. – The range of sequence numbers must be increased. – The sender and receiver sides of the protocols may have to buffer more than one packet. – Go-Back-N and Selective Repeat approaches Pipelined reliable data transfer protocols

23 o No more than some max allowable number, N, of unacknowledged packets in the pipeline o Base: the seq number of the oldest unacknowledged packet o Nextseqnum: the smallest unused sequence number (= the seq number of the next packet to be sent) o Fig 3.19 o [0, base-1], [base, nextseqnum-1], [nextseqnum, base+N-1], [base+N, ] o Window size, sliding-window protocol Go-Back-N (GBN)

24 o EFSM of GBN sender – Invocation from above – Cumulative acknowledgement: all packets with a seq number up to and including n have been correctly received at the receiver. – A timeout event for the oldest transmitted but not yet acknowledged packet. o EFSM of GBN sender – The receiver discards out-of-order packets – The receiver need not buffer any out-of-order packets.

25 o When the window size and bandwidth-delay product are both large, many packets can be in the pipeline. o The sender retransmit only those packets that it suspects were received in error (were lost or corrupted) o The receiver individually acknowledge correctly received packets. o The sender will have already received ACKs for some of the packets in the window. o Synchronization between sender and receiver windows -> the window size must be less than or equal to half the size of the seq number space for SR protocols. Selective Repeat (SR)

26 o Synchronization between sender and receiver windows -> the window size must be less than or equal to half the size of the seq number space for SR protocols.