Chapter 6: Network Communications and Protocols

1
Chapter 6Network Communications and Protocols

2
Learning Objectives

• Understand the function and structure of packets
  in a network, and analyze and understand those
  packets
• Understand the function of protocols in a network
• Discuss the layered architecture of protocols,
  and describe common protocols and their
  implementation
• Understand channel access methods

3
Function of Packets in Network Communications

• Networks reformat data into smaller, more
  manageable pieces called packets or frames
• Advantages of splitting data include
• More efficient transmission, since large units of
  data saturate network
• More computers able to use network
• Faster transmissions since only packets
  containing errors need to be retransmitted

4
Packet Structure

• Three basic parts of packet, as seen in Header
  contains source and destination address along
  with clocking information to synchronize
  transmission
• Data payload or actual data can vary from 512
  bytes to 16 kilobytes
• Trailer information to verify packets
  contents, such as Cyclic Redundancy Check (CRC)

5
Typical Packet Structure
6
Packet Size
7
Packet Creation

• From sender, data moves down layers ofOSI model
• Each layer adds header or trailer information
• Data travels up layers at receiver
• Each layer removes header or trailer information
  placed by corresponding sender layer

8
Header/Trailer Information Added or Removed
9
Packet Creation (continued)

• Outgoing data stream enters OSI model as complete
  message
• Remains as data at layers 5-7
• Lower layers split data
• Transport layer 4 splits it into segments
• Network layer 3 splits segments into packets
• Data Link layer 2 puts packets into frames
• Physical layer 1 transmits packets as bits

10
Understanding Packets

• Three kinds of packets
• Unicast packet addressed to only one computer
• Broadcast packet created for all computers on
  network
• Multicast packet created for any computers on
  network that listen to shared network address

11
Protocols

• Rules and procedures for communicating
• To communicate, computers must agree on
  protocols
• Many kinds of protocols
• Connectionless
• Connection-oriented
• Routable
• Nonroutable

12
The Function of Protocols

• Each protocol has different purpose and function
• Protocols may work at one or more layers
• More sophisticated protocols operate at higher
  layers of OSI model
• Protocol stack or protocol suite is set of
  protocols that work cooperatively
• Most common protocol stack is TCP/IP used by the
  Internet and pretty much all operating systems

13
Protocol Functions

• Fragmentation and reassembly
• Flow control
• Error control

14
Fragmentation and Reassembly

• Exchange data between two entities
• Characterized as sequence of PDUs of some bounded
  size
• Application level message
• Lower-level protocols may need to break data up
  into smaller blocks
• Communications network may only accept blocks of
  up to a certain size
• ATM 53 octets
• Ethernet 1526 octets

15
Disadvantages of Fragmentation

• Make PDUs as large as possible because
•  PDU contains some control information
• Smaller block, larger overhead
• PDU arrival generates interrupt
• Smaller blocks, more interrupts
• More time processing smaller, more numerous PDUs
•  

16
Reassembly

• Segmented data must be reassembled into messages
• More complex if PDUs out of order

17
PDUS and Fragmentation
18
Flow Control

• Performed by receiving entity to limit amount or
  rate of data sent
• Stop-and-wait
• Each PDU must be acknowledged before next sent
• Credit
• Amount of data that can be sent without
  acknowledgment
• E.g. HDLC sliding-window
• Must be implemented in several protocols
• Network traffic control
• Buffer space
• Application overflow
• E.g. waiting for disk access

19
Error Control

• Guard against loss or damage
• Error detection and retransmission
• Sender inserts error-detecting code in PDU
• Function of other bits in PDU
• Receiver checks code on incoming PDU
• If error, discard
• If transmitter doesnt get acknowledgment in
  reasonable time, retransmit
• Error-correction code
• Enables receiver to detect and possibly correct
  errors
• Error control is performed at various layers of
  protocol
• Between station and network
• Inside network

20
Design Issues

• Routing
• Datagram lifetime
• Fragmentation and re-assembly
• Error control
• Flow control

21
Routing

• End systems and routers maintain routing tables
• Indicate next router to which datagram should be
  sent
• Static
• May contain alternative routes
• Dynamic
• Flexible response to congestion and errors
• Source routing
• Source specifies route as sequential list of
  routers to be followed
• Security
• Priority
• Route recording

22
Datagram Lifetime

• Datagrams could loop indefinitely
• Consumes resources
• Transport protocol may need upper bound on
  datagram life
• Datagram marked with lifetime
• Time To Live field in IP
• Once lifetime expires, datagram discarded (not
  forwarded)
• Hop count
• Decrement time to live on passing through a each
  router
• Time count
• Need to know how long since last router

23
Fragmentation and Re-assembly

• Different packet sizes
• When to re-assemble
• At destination
• Results in packets getting smaller as data
  traverses internet
• Intermediate re-assembly
• Need large buffers at routers
• Buffers may fill with fragments
• All fragments must go through same router
• Inhibits dynamic routing

24
IP Fragmentation (1)

• IP re-assembles at destination only
• Uses fields in header
• Data Unit Identifier (ID)
• Identifies end system originated datagram
• Source and destination address
• Protocol layer generating data (e.g. TCP)
• Identification supplied by that layer
• Data length
• Length of user data in octets

25
IP Fragmentation (2)

• Offset
• Position of fragment of user data in original
  datagram
• In multiples of 64 bits (8 octets)
• More flag
• Indicates that this is not the last fragment

26
Fragmentation Example
27
Fragmentation Header

• Fragmentation only allowed at source
• No fragmentation at intermediate routers
• Node must perform path discovery to find smallest
  MTU(max transmission unit) of intermediate
  networks
• Source fragments to match MTU
• Otherwise limit to 1280 octets

28
Fragmentation Header Fields

• Next Header
• Reserved
• Fragmentation offset
• Reserved
• More flag
• Identification

29
Error Control

• Not guaranteed delivery
• Router should attempt to inform source if packet
  discarded
• e.g. for time to live expiring
• Source may modify transmission strategy
• May inform high layer protocol
• Datagram identification needed

30
Flow Control

• Allows routers and/or stations to limit rate of
  incoming data
• Limited in connectionless systems
• Send flow control packets
• Requesting reduced flow
• e.g. ICMP

31
Protocols in a Layered Architecture

• Most protocols can be positioned and explained in
  terms of layers of OSI model
• Protocol stacks may have different protocols for
  each layer
• See Figure 6-3 for review of functions of each
  layer of OSI model
• See Figure 6-4 for three major protocol types
• Application protocols at layers 5-7
• Transport protocols at layer 4
• Network protocols at layers 1-3

32
Functions of OSI Model Layers
33
Three Main Protocol Types
34
Network Protocols

• Provide addressing and routing information, error
  checking, and retransmission requests
• Services provided by network protocols are called
  link services
• Popular network protocols include
• Internet Protocol version 4 (IPv4)
• Internetwork Packet Exchange (IPX) and NWLink
• NetBEUI
• Internet Protocol version 6 (IPv6)

35
Transport Protocols

• Handle data delivery between computers
• May be connectionless or connection-oriented
• Transport protocols include
• Transmission Control Protocol (TCP)
• Sequenced Packet Exchange (SPX) and NWLink
• NetBIOS/NetBEUI

36
Application Protocols

• Operate at upper layers of OSI model to provide
  application-to-application service
• Some common application protocols are
• Simple Mail Transport Protocol (SMTP)
• File Transfer Protocol (FTP)
• Simple Network Management Protocol (SNMP)
• NetWare Core Protocol (NCP)
• AppleTalk File Protocol (AFP)

37
Common Protocol Suites

• Combination of protocols that work
  cooperatively to accomplish network
  communications
• Some of the most common protocol suites are

• TCP/IP
• NWLink (IPX/SPX)
• NetBIOS/NetBEUI
• AppleTalk

• DLC
• XNS
• DECNet
• X.25

38
Transmission Control Protocol/ Internet Protocol
(TCP/IP)

• Called the Internet Protocol (IP)
• Most commonly used protocol suite for networking
• Excellent scalability and superior functionality
• Able to connect different types of computers and
  networks
• Default protocol for Novell NetWare, Windows
  XP/2000/2003, all Unix/Linux varieties, and Mac
  OS X
• See Figure 6-5 for relationship to OSI model

39
TCP/IP Compared to OSI Model
40
IP Addressing

• Logical addresses, 32-bits or 4 bytes long
• Four octets separated by periods, each with
  decimal value from 0-255
• First part of address identifies network
• Second part of address identifies host or
  individual computer
• IP addresses broken into classes
• Number of IP address registries under control of
  Internet Assigned Numbers Authority (IANA)

41
Classless Inter-Domain Routing (CIDR)

• Internet uses CIDR
• Demarcation between network and host not always
  based on octet boundaries
• May be based on specific number of bits from
  beginning of address
• Called subnetting, the process involves
  stealing bits from host portion of address for
  use in network address
• Provides fewer hosts on each network but more
  networks overall

42
IPv4 Header
43
Header Fields (1)

• Version
• Currently 4
• IP v6 - see later
• Internet header length
• In 32 bit words
• Including options
• Type of service
• Total length
• Of datagram, in octets

44
Header Fields (2)

• Identification
• Sequence number
• Used with addresses and user protocol to identify
  datagram uniquely
• Flags
• More bit
• Dont fragment
• Fragmentation offset
• Time to live
• Protocol
• Next higher layer to receive data field at
  destination

45
Header Fields (3)

• Header checksum
• Reverified and recomputed at each router
• 16 bit ones complement sum of all 16 bit words in
  header
• Set to zero during calculation
• Source address
• Destination address
• Options
• Padding
• To fill to multiple of 32 bits long

46
Data Field

• Carries user data from next layer up
• Integer multiple of 8 bits long (octet)
• Max length of datagram (header plus data) 65,535
  octets

47
IPv4 Address Formats
48
IP Addresses - Class A

• 32 bit global internet address
• Network part and host part
• Class A
• Start with binary 0
• All 0 reserved
• 01111111 (127) reserved for loopback
• Range 1.x.x.x to 126.x.x.x
• All allocated

49
IP Addresses - Class B

• Start 10
• Range 128.x.x.x to 191.x.x.x
• Second Octet also included in network address
• 214 16,384 class B addresses
• All allocated

50
IP Addresses - Class C

• Start 110
• Range 192.x.x.x to 223.x.x.x
• Second and third octet also part of network
  address
• 221 2,097,152 addresses
• Nearly all allocated
• See IPv6

51
Routing Using Subnets
52
ICMP

• Internet Control Message Protocol
• RFC 792
• Transfer of (control) messages from routers and
  hosts to hosts
• Feedback about problems
• e.g. time to live expired
• Encapsulated in IP datagram
• Not reliable

53
ICMP Message Formats
54
IP v6 - Version Number

• IP v 1-3 defined and replaced
• IP v4 - current version
• IP v5 - streams protocol
• IP v6 - replacement for IP v4
• During development it was called IPng
• Next Generation

55
IPv6

• Current four byte version is IPv4
• Now reaching limit of 4-byte addresses
• IPv6 being used now on the Internet backbone and
  other large networks
• Uses 16 byte (128-bit) addresses
• Retains backward compatibility with IPv4 4-byte
  addresses
• Will provide limitless supply of addresses

56
Why Change IP?

• Address space exhaustion
• Two level addressing (network and host) wastes
  space
• Network addresses used even if not connected to
  Internet
• Growth of networks and the Internet
• Extended use of TCP/IP
• Single address per host
• Requirements for new types of service

57
IPv6 Enhancements (1)

• Expanded address space
• 128 bit
• Improved option mechanism
• Separate optional headers between IPv6 header and
  transport layer header
• Most are not examined by intermediate routes
• Improved speed and simplified router processing
• Easier to extend options
• Address autoconfiguration
• Dynamic assignment of addresses

58
IPv6 Enhancements (2)

• Increased addressing flexibility
• Anycast - delivered to one of a set of nodes
• Improved scalability of multicast addresses
• Support for resource allocation
• Replaces type of service
• Labeling of packets to particular traffic flow
• Allows special handling
• e.g. real time video

59
IPv6Structure
60
IP v6 Header
61
IP v6 Header Fields (1)

• Version
• 6
• Traffic Class
• Classes or priorities of packet
• Still under development
• See RFC 2460
• Flow Label
• Used by hosts requesting special handling
• Payload length
• Includes all extension headers plus user data

62
IP v6 Header Fields (2)

• Next Header
• Identifies type of header
• Extension or next layer up
• Source Address
• Destination address

63
IPv6 Addresses

• 128 bits long
• Assigned to interface
• Single interface may have multiple unicast
  addresses
• Three types of address

64
Subnet Masks

• Part of IP address identifies network and part
  identifies host
• IP uses subnet mask to determine what part of
  address identifies network and what part
  identifies host
• Network section identified by binary 1
• Host section identified by binary 0

65
Network Address Translation (NAT)

• Allows organization to use private IP addresses
  while connected to the Internet
• Performed by network device such as router that
  connects to Internet
• See Simulation 6-3 and Figure 6-6 for examples of
  NAT

66
Network Address Translation (NAT) (continued)
67
Dynamic Host Configuration Protocol (DHCP)

• DHCP server receives block of available IP
  addresses and their subnet masks
• When computer needs address, DHCP server selects
  one from pool of available addresses
• Address is leased to computer for designated
  length and may be renewed
• Can move computers with ease no need to
  reconfigure IP addresses
• Some systems, such as Web servers, must have
  static IP address

68
NetBIOS and NetBEUI

• Consortium of Microsoft, 3Com, and IBM developed
  lower-level protocol NetBEUI in mid-1980s
• NetBIOS Extended User Interface
• Spans layers 2, 3, and 4 of OSI model
• Both designed for small- to medium-sized
  networks, from 2-250 computers

69
NetBIOS and NetBEUI (continued)

• Figure 6-7 shows Microsoft protocol suite and its
  relationship to OSI model
• Defines four components above Data Link layer
• Runs on any network card or physical medium
• Redirector interprets requests and determines
  whether they are local or remote
• If remote, passes request to Server Message Block
  (SMB)
• SMB passes information between networked computers

70
Microsoft Protocol Suite Compared to OSI Model
71
NetBIOS and NetBEUI (continued)

• NetBEUI works at Transport layer to manage
  communications between two computers
• Nonroutable protocol skips Network layer
• NetBEUI packet does not contain source or
  destination network information

72
NetBIOS and NetBEUI (continued)

• NetBIOS operates at Session layer to provide
  peer-to-peer network application support
• Unique 15-character name identifies each computer
  in NetBIOS network
• NetBIOS broadcast advertises computers name
• Connection-oriented protocol, but can also use
  connectionless communications
• Nonroutable protocol, but can be routed when
  using routable protocol for transport

73
NetBIOS and NetBEUI (continued)

• NetBEUI is small, fast, nonroutable Transport and
  Data Link protocol
• All Windows versions include it
• Ideal for DOS based computers
• Good for slow serial links
• Limited to small networks
• Server Message Block operates at Presentation
  layer
• Used to communicate between redirector and server
  software

74
IPX/SPX

• Original protocol suite designed for Novells
  NetWare network operating system
• Still supported with NetWare 6.0, but TCP/IP is
  now primary protocol
• NWLink is Microsofts implementation of IPX/SPX
  protocol suite
• Figure 6-8 shows protocols in NWLink and
  corresponding OSI layers
• Must consider which Ethernet frame type with
  NWLink

75
NWLink Compared to OSI Model
76
AppleTalk

• Defines physical transport in Apple Macintosh
  networks
• Divides computers in zones
• AppleTalk Phase II allows connectivity outside
  Macintosh world

77
Implementing and Removing Protocols

• Easy to add or remove protocols
• TCP/IP loads automatically when most operating
  systems are installed
• In Windows 2000/2003/XP, use Local Area
  Connections Properties to add or remove protocols
• See Figure 6-9

78
Network and Dial-up Connections
79
Putting Data on the Cable Access Methods

• Consider several factors
• How computers put data on the cable
• How computers ensure data reaches destination
  undamaged

80
Function of Access Methods

• Rules specify when computers can access cable or
  data channel
• Channel access methods assure data reaches its
  destination
• Prevents two or more computers from sending
  messages that may collide on cable
• Allows only one computer at a time to send data

81
Major Access Methods

• Channel access is handled at Media Access Control
  (MAC) sublayer of Data Link layer
• Five major access methods
• Contention
• Switching
• Token passing
• Demand priority
• Polling

82
Contention

• In early networks, contention method allowed
  computers to send data whenever they had data to
  send, resulting in frequent collisions and
  retransmissions
• Figure 6-11 shows data collision
• Two carrier access methods were developed for
  contention-based networks
• Carrier Sense Multiple Access with Collision
  Detection (CSMA/CD)
• Carrier Sense Multiple Access with Collision
  Avoidance (CSMA/CA)

83
Data Collision
84
CSMA/CD

• Popular access method used by Ethernet
• Prevents collisions by listening to channel
• If no data on line, may send message
• If collision occurs, stations wait random period
  of time before resending data
• See Figure 6-11

85
CSMA/CD (continued)
86
CSMA/CD (continued)

• Limitations and disadvantages of CSMA/CD
• Not effective at distances over 2500 meters
• More computers on network likely to cause more
  collisions
• Computers have unequal access to media
• Computer with large amount of data can monopolize
  channel

87
CSMA/CA

• Uses collision avoidance, rather than detection,
  to avoid collisions
• When computer senses channel is free, it signals
  its intent to transmit data
• Used with Apples LocalTalk
• Advantages and disadvantages
• More reliable than CSMA/CD at avoiding collisions
• Intent to transmit packets add overhead and
  reduce network speed

88
Switching

• Switch interconnects individual nodes and
  controls access to media
• Switching usually avoids contention and allows
  connections to use entire bandwidth
• Other advantages include
• Fairer than contention-based technology
• Permits multiple simultaneous conversations
• Supports centralized management
• Disadvantage include
• Higher cost
• Failure of switch brings down network

89
Token Passing

• Token passes sequentially from one computer to
  next
• Only computer with token can send data, as seen
  in Figure 6-12
• Advantages and disadvantages
• Prevents collisions
• Provides all computers equal access to media
• Computer must wait for token to transmit, even if
  no other computer wants to transmit
• Complicated process requires more expensive
  equipment

90
Communication in a Token-Passing Network
91
Demand Priority

• Used only by 100VG-AnyLAN 100 Mbps Ethernet
  standard (IEEE 802.12)
• Runs on star bus topology, as seen in Figure 6-13
• Intelligent hubs control access to network
• Computer sends hub demand signal when it wants to
  transmit
• Advantages and disadvantages
• Allows certain computers to have higher
  priorities
• Eliminates extraneous traffic by not broadcasting
  packets but sending them to each computer
• Price is major disadvantage

92
Demand Priority Uses Star Bus Topology
93
Polling

• One of oldest access methods
• Central controller, called primary device, asks
  each computer or secondary device if it has data
  to send, as seen in Figure 6-14
• Advantages and disadvantages
• Allows all computers equal access to channel
• Can grant priority for some computers
• Does not make efficient use of media
• If primary device fails, network fails

94
Primary Device Controls Polling
95
Choosing an Access Method

• Network topology is biggest factor in choosing
  access method
• Ring topology usually uses token-passing
• Switching can emulate all common topologies