Chapter 16' Protocols And Layering

1
Chapter 16. Protocols And Layering

• Jing Wang
• Towson University

2
Protocols

• Communication protocol
• An agreement that specifies the format and
  meaning of messages computers exchange
• An application interacts with protocol software
  that follows the rules of a given protocol.
• Suites (families)
• Protocols are designed and developed in complete,
  cooperate sets

3
A Plan For Protocol Design

• 7-layer Reference Model
• Defined by International Organization for
  Standardization (ISO)
• Replaced by Internet Reference Model

4

• Figure 16.1. The historic ISO 7-Layer Reference
  Model. A layering model is a tool to help
  protocol designers construct a suite of protocols
  that solves all communication problems.

5
The Seven Layers

• Layer 1 Physical
• Basic network hardware.
• e.g. specification of RS232, detailed
  specification of LAN hardware
• Layer 2 Data Link
• How to organize data into frames, how to transmit
  frames over a network
• e.g. frame format, bit or byte stuffing, checksum
  computation
• Layer 3 Network
• Specify how addresses are assigned, how packets
  are forwarded from one end of the network to
  another
• e.g. specification of addressing, details of
  forwarding
• Layer 4 Transport
• Specify how to handle details of reliable transfer

6
The Seven Layers

• Layer 5 Session
• Specify how to establish a communication session
  with a remote system
• e.g. how to login to a remote timesharing
  computer
• e.g. specification for security details such as
  authentication using passwords
• Layer 6 Presentation
• Specify how to represent data. (because different
  brands of computers use different internal
  representations for integers and characters)
• Layer 7 Application
• Specifies how one particular application uses a
  network
• e.g. specification for an application that
  transfers files

7
Stacks Layered Software

• Figure 16.2. The conceptual path of data as it
  travels from an application on computer 1 across
  a network to an application on computer 2.

8
Multiple, Nested Headers

• Figure 16.4. The nested protocol headers that
  appear in a frame as the frame travels across a
  network if the full ISO stack is used. Each layer
  of protocol software adds a header to an outgoing
  frame.

9
The Scientific Basis For Layering

• Layer N software on the destination computer must
  receive the exact message sent by layer N
  software on the sending computer

10

• Figure 16.5. The layering principle applied at
  each layer of the old ISO model. If protocol
  software on the sending computer changes the
  message, the change must be reversed by the
  corresponding protocol software on the receiver.

11
Techniques Protocols Use

• Techniques that data link protocol use to detect
  bits corruption during transmission
• Parity bit, frame checksum, CRC
• Techniques protocols use
• Sequencing for out-of-order delivery
• Sequencing to eliminate duplicate packets
• Retransmitting lost packets
• Avoiding replay caused by excessive delay
• Flow control to prevent data overrun
• Mechanism to avoid network congestion

12
Techniques Protocols Use

• Retransmitting lost packets
• To guarantee reliable transfer (i.e., transfer
  without loss), protocols use positive
  acknowledgement with retransmission
• Acknowledgement (ACK)
• timer

13
Techniques Protocols Use

• Avoiding replay caused by excessive delay
• Packets from different conversation

14
Techniques Protocols Use

• Flow control to prevent data overrun
• Stop-and-go
• Sliding window
• A fixed window
• Maximum amount of data that can be sent before an
  acknowledgement arrives
• Tw Tg x W
• Tw throughput of sliding window protocol
• Tg throughput of stop-and-go protocol
• W window size
• Tw min(B, Tg x W)
• B underlying hardware bandwidth

15

• Figure 16.6. A 4-packet window sliding through
  outgoing data. The window is shown (a) when
  transmission begins, (b) after two packets have
  been acknowledged, and (c) after eight packets
  have been acknowledged. The sender can transmit
  all packets in the window.

16

• Figure 16.7. Messages required to send a sequence
  of four packets using (a) stop-and-go flow
  control, and (b) a 4-packet sliding window. Time
  proceeds down the page, and each arrow shows one
  message sent from one computer to the other.

17
Reliable Transmission

• Two fundamental mechanisms
• Acknowledgements (ACK)
• Timeouts
• General strategy to implement
• Automatic repeat request (ARQ)
• ARQ algorithms
• Stop-and wait
• Sliding window
• Reliable delivery Reliably deliver frames across
  an unreliable link
• Ordered delivery Preserve the order in which
  frames are transmitted
• Flow control Support flow control (keep the send
  from overrunning the receiver)
• Bandwidth x delay product
• Represents the amount of data that could be in
  transit
• Principle keeping the pipe full

18
Figure 2.21. Timeline for the sliding window
algorithm
19
The sliding window algorithm

• SeqNum sequence number

20
LFS LAR SWS SWS send window size LAR last
acknowledgement received LFS last frame sent
Figure 2.22. Sliding window on sender
21
LAF-LFR RWS RWS receive window size LAF
largest acceptable frame LFR last frame received
Figure 2.23. Sliding window on receiver
22
TCPs variant of the sliding window

• Purposes
• Guarantees the reliable delivery of data
• Ensures that data is delivered in order
• Enforces flow control between the sender and the
  receiver
• Difference in flow control
• Receiver advertises a window size to the sender
• AdvertisedWindow field in the TCP header
• Based on the amount of memory allocated to the
  connection for the purpose of buffering data

23
Figure 5.8. Relationship between TCP send buffer
(a) and receive buffer (b)
24
Techniques Protocols Use

• Mechanism to avoid network congestion
• Congestion
• More packets arrive than can be sent, the queue
  grows and the effective delay increases.
• Congestion collapse
• The situation persists, the entire network become
  unusable
• Avoid congestion collapse
• Arrange for packet switches to inform senders
  when congestion occurs
• Use packet loss as an estimate of congestion
• Modern network hardware works well most packet
  loss results from congestion, not hardware
  failure
• Packet loss can be measured if the sends uses a
  timeout and retransmission strategy
• Rate reduction

25
16.12. Summary

• Protocol software
• Layering principle
• Protocol stack
• Techniques protocols use
• Sequencing to handle out-of-order and duplicate
  packets
• Acknowledgement and retransmission to handle
  packet loss
• Unique session IDs to prevent replay
• A stop-and-go or sliding window mechanism to
  control the flow of data
• Rate reduction to handle network congestion