Wide Area Networks

1
Wide Area Networks

• ICS31 Computer Networks Data Communications
• Topic 6

2

• Wide Area Network Concepts

3
Section Preview
In this section we will study

• WAN terminology and topology
• Functions of the data link layer
• WAN data link protocols
• Functions of the network layer
• Network routing
• Differences between LANs and WANs

4
WAN Terminology

• Link
• The direct connection of two nodes in a network.
  A link, therefore, connect two computers.
• End-to-End Routing
• Sometimes, when one node wants to send a message
  to another node, the two are not directly
  connected. The message must then pass through one
  or more intermediate nodes before arriving at the
  final destination. Determining how this is done
  is called end-to-end routing.
• Path
• Th links that the message traverses.
• Hop
• The number of hops a message takes in going from
  its source to its destination is the number of
  links it traverses.

5
WAN Terminology (cont.)

• Store-and-Forward
• A technique used by some networks to send data
  along a path. Each intermediary node along the
  path stores the message, sends an acknowledgment
  of message receipt to the sender, and then
  forwards the message to the next node on the
  path. When the sender receives an acknowledgement
  that the message bas been received by the next
  node, it is no longer responsible for
  retransmitting the message if an error occurs.
• Session
• A communication between two users of a network. A
  user can be a terminal operator, an application,
  or any other originator of messages. In some
  systems, sessions are quite formal, with
  well-defined conventions for establishing,
  continuing, and terminating the dialogue.
• Packet Switching
• The technology of transmitting a message in one
  or more fixed-length data packets. (Contrast this
  with circuit switching systems such as the
  telephone network.)

6
Hierarchical Network

• The hierarchical network topology is also called
  a tree structure.
• This type of network closely resembles corporate
  organization charts, and corporate computer
  centers are one place in which this topology can
  be found.
• Information flowing from a district in one
  division to a district in a different division
  would need to go through the root or corporate
  node. This topology allow for a great deal of
  network control.

7
Interconnected Network

• This topology provides a high degree of
  flexibility because many paths are available
  between nodes if a link should fail, so
  congestion can be avoided.
• Costs can also be controlled because
  interconnected topology is capable of the
  shortest or least expensive configuration.

8
Combination Networks

• Hierarchical and interconnected networks can be
  integrated into one network.
• One such combination is a backbone networkfor
  instance, a ringwith spurs attached.
• In widely distributed systems with a large number
  of nodes, this configuration helps reduce the
  number of hops, the length of the path, and
  congestion problems.
• Backbone networks are appearing more often in LAN
  technology, as well.

9
Hierarchical Network Topology
Corporate Headquarters Computer
Regional Office Computers
District Office Computers
Branch Office Computers
10
Interconnected Network Topology
11
Backbone Network Topology
12
Bit Synchronous Protocols

• Synchronous data link control (SDLC) from IBM
• Advanced data communications control procedure
  (ADCCP), an ANSI standard data link protocol
• High-level data link control (HDLC), a standard
  of the International Standards Organization (ISO)
• Link access procedurebalanced (LAPB), designated
  as the data link protocol for X.25 packet
  distribution networks.

13
HDLC Frame Format
Flag 01111110
Address (1 or more octets)
Control Field (8 or 16 bits)
Data (Optional Octets)
Frame Check Sequence 16 or 32 bits)
Flag 01111110
14
Expansion of Control Field in HDLC Frame
Ns (3 bits)
P/F Bit
Nr (3 bits)
0
8-Bit Control Field
Ns (7 bits)
P/F Bit
Nr (7bits)
0
16-Bit Control Field
15
Message Routing in a WAN

• Centralized Routing
• In centralized routing, one node is designated as
  the network routing manager to whom all nodes
  periodically forward such status information as
  queue lengths on outgoing and incoming lines and
  the number of messages processed within the most
  recent time interval.
• Increased congestion and large processing power
  requirements are among the disadvantages of
  centralized routing.
• Distributed Routing
• Distributed routing relies on each node to
  calculate its own best routing table, which
  requires each node to periodically transmit its
  status to its neighbors.
• Static Routing
• The purest form of static routing involves always
  using one particular path between two nodes if a
  link in that path is down, then communication
  between those nodes is impossible.

16
Message Routing in a WAN (cont.)

• Weighted Routing
• When multiple paths exist, some implementations
  use weighted routing, in which each path is
  weighted according to perceived use. The path is
  then randomly selected from the weighted
  alternatives.
• Adaptive Routing
• Adaptive routing, occasionally called dynamic
  routing, attempts to select the quickest or best
  current route for the message or session.
• Broadcast Routing
• Broadcast routing is exemplified by CSMA/CD and
  token passing. The message is broadcast to all
  stations, and only the station to which the
  message is addressed accepts it.

17
Centralized Routing
Old Path
New Path
18
Weighted Routing
20
50
30
19
Adaptive Routing
Very Busy
Idle
Idle
Congestion Avoidance Path
Idle
20
Packet Distribution Networks

• PDN Terminology
• A packet distribution network (PDN) is a
  packet-switching WAN variously called an X.25, a
  value-added network (VAN), or a public data
  network. The terms packet distribution and packet
  switching both refer to how data are transmitted
  (that is, as one or more packets with a fixed
  length).
• PDN Advantages and Disadvantages
• The user is charge for the amount of data
  transmitted rather than for connect time. The PDN
  gives access to many different locations without
  the cost of switched connections. Access to the
  PDN is usually via a local telephone call, which
  also reduces costs.
• But, because the PDN is usually shared, users
  must compete with each other for circuits. If the
  number of data packets to be transferred is
  great, then the cost of using a PDN can exceed
  that of leased facilities.
• Frame Relay and Asynchronous Transfer Mode (ATM)
• The overhead of the X.25 protocol results in the
  transmission speed being limited to 64 Kbps or
  less. Consequently, X.25 is not well suited to
  some forms of transmission such as voice and
  video, where frames must arrive closely together.
  This motivated the development of newer packet
  switching technologies such as frame relay and
  ATM).

21
Comparing WANs with LANs

• The primary difference between a WAN and a LAN is
  distance. A LAN serves a limited geographical
  are, typically within one building or building
  complex. A WAN can cover a large geographical
  area.
• Topology, Protocols, and Routing
• LAN topologies are usually bus, ring or star.
  WAN topologies are typically hierarchical or
  interconnected, although rings and stars are also
  used.
• Media
• LAN media are usually twisted-pair wires, coaxial
  cable, fiber optic cable, or one of the new
  wireless media. A single LAN usually uses a
  single medium type. WAN media are often obtained
  through a common carrier and may consist of a
  variety of media, such as telephone wires, fiber
  optic cable, coaxial cable, microwave radio, and
  satellite.
• Ownership
• A LAN is almost always privately controlled with
  respect to hardware, software, and media. A WAN,
  on the other hand, usually consists of computer
  hardware and software owned or controlled by the
  user, together with media and associated data
  communication equipment provided by a common
  carrier.

22
Comparing WANs with LANs (cont.)

• Transmission Speed
• LAN nodes are connected via high-speed
  communication path. The speed typically is at
  least 1 Mbps and often higher. Transmission
  speeds for WANs vary widely and, like those of
  LANs, are constantly increasing. Common WAN
  speeds are 9600 bps, 56 Kbps, and 1.54 Mbps.
  Speeds of 1.54 Mbps result from a transmission
  service known as T-1.

23
LAN-WAN Interconnections

• The differences between LANs and WANs just cited
  present considerable obstacles to LAN-WAN
  interconnection. When we connect a LAN to a WAN,
  many of these differences must be reconciled. LAN
  speeds can be 10,000 times faster than WAN
  speeds. The format of a LAN message can be
  different from that of the WAN to which it is
  connected, and the interconnection must translate
  messages from one protocol format to the other.

24
Making Network Connections
25
Section Preview
In this section we will study

• LAN-to-LAN connections
• Repeaters, bridges, routers, and gateways
• Switches and virtual LANs
• Remote access
• Common carrier services

26
LAN-to-LAN Connections

• Companies that have LANs in separate geographical
  locations or LANs that cover distances greater
  than the maximum medium distance allowed, or
  companies with more nodes that can be
  accommodated by one LAN, must segment their
  network into two or more LANs.
• Having separate LANs allows a company to split
  functions and gives an additional level of
  security.
• LANs may also be connected in order to
  consolidate independent LANs that may have been
  formed in an ad-hoc manner.
• Responsiveness of the system can be maintained
  while the number of users increases by adding
  more resources to an existing LANmemory, disks,
  another server, etc. or by splitting the LAN
  into two or more smaller LANs.

27
Repeaters

• Every LAN has a distance restriction. IEEE
  standards specify a maximum segment length of 500
  meters. If you want to span longer distances, you
  can use a repeater to connect two segments.
• As signals travel along the medium, they lose
  strength through attenuation. Weak signals can
  cause transmission errors. A repeater accepts a
  signal, regenerates it, and passes it along at
  full strength.
• A repeater does not separate one segment of the
  network from another.

28
A Repeater, Bridge, and Router and the OSI
Reference Model
Processor 1
Processor 2
Application Layer
Application Layer
Presentation Layer
Presentation Layer
Session Layer
Session Layer
Transport Layer
Transport Layer
Router
Network Layer
Network Layer
Data Link Layer
Data Link Layer
Bridge
Repeater
Physical Layer
Physical Layer
29
Bridges

• Early bridges were used to connect two networks,
  each of which used the same MAC protocol. Today,
  bridges also connect LANs having different MAC
  protocols. This device may be called a brouter
  or multiprotocol bridge.
• Most bridges being sold today are called learning
  bridges, or transparent bridges. A learning
  bridge builds its routing table from messages it
  receives and does not need to be loaded with a
  predefined routing table.

30
Basic Bridge Functions
Packet Routing Function Additional
Functions
1. Accept packet from LAN A. 2. Examine address
of packet. 3. If packet address is a LAN A
address, allow the packet to continue on LAN
A. 4. If packet address is a LAN B address,
transmit the packet onto the LAN B medium. 5. Do
the equivalent for LAN B packets. Media
conversion Learning Remote connection Signal
conversion Speed conversion Packet
statistics Token ring to ethernet conversion
31
Token Rings Connected by a Bridge
LAN A Token-Ring Network
Bridge
LAN B Token-Ring Network
32
Spanning Tree Algorithms

• Spanning tree algorithms, in which bridges
  exchange routing information with each other, can
  be used on any type of LAN.
• The advantages of the spanning tree algorithm are
  that it is MAC-layer-independent, bridges can
  learn the topology of the network without manual
  intervention, and paths can change if an existing
  path becomes inoperable or if a better path is
  introduced. The algorithm overhead is the size of
  the routing table for networks with many
  communicating nodes, and the extra network
  traffic resulting from status messages and
  flooding.

33
Source Routing

• Source routing is an IEEE standard used as a
  routing algorithm for token-passing networks.
• The advantage of the source routing algorithm is
  that bridges are not responsible for maintaining
  large routing tables for extensive networks. Each
  node is responsible for maintaining routing
  information only for the nodes with which it
  communicates. The disadvantages are the overhead
  of sending numerous packets during discovery and
  the extra routing data that must be appended to
  each message.

34
Remote Bridge Connection Alternatives
RS-232 serial lines Synchronous transmission at
56 Kbps or 64 Kbps Fractional T-1 at multiples of
64 Kbps Integrated services digital network
(ISDN) Digital subscriber lines (DSL) RS-422
serial lines at 19.2 Kbps to 2 Mbps T-1 Line at
1.5 Mbps X.25 packet-switching network Frame
relay Asynchronous Transfer Mode (ATM)
35
Routers

• Routers operate at the network layer of the OSI
  Reference Model. The network layer is responsible
  for packet routing and for collecting accounting
  information.
• Some networks use a static routing algorithm,
  meaning that packet routing between two nodes
  never changes. In a CSMA/CD bus LAN, a packet is
  broadcast to every node in a token ring, a
  packet is transmitted from one node to the next
  node in the ring.
• A Novell network uses a protocol called sequenced
  packet exchange/internetwork packet exchange
  (SPX/IPX) to transfer packets between nodes. SPX
  operates at the transport layer and IPX at the
  network layer.
• Another protocol used by many networks is the
  transmission control protocol/internet protocol
  (TCP/IP). TCP/IP is the protocol used on the
  Internet.

36
Gateways

• The interface between two dissimilar networks is
  called a gateway. A gateway is basically a
  protocol converter. It reconciles the differences
  between the networks it connects.
• The components of a gateway are the network
  interfaces and the logic that carries out the
  conversion necessary when moving messages between
  networks. The conversion must change the header
  and trailer of the packet to make it consistent
  with the protocol of the network or data link to
  which the message is being transferred. (In some
  cases considered as a device for connecting two
  networks with different application layer
  protocols.)

37
Ways to Increase LAN Media Throughput

• Faster hardware
• Higher network speeds
• Lower bandwidth demands
• LAN segmentation
• Full duplex
• Switching hubs

38
LAN Switches

• LAN switches work similarly to the way in which a
  telephone switch works. In the idle state, you
  are not connected to anything except the
  telephone switch and cannot communicate with
  anyone until a circuit is set up. When you dial
  someones number, the complete transmission
  capacity of the telephone network is not
  dedicated to your call instead, a connection is
  made between your telephone and the telephone of
  the person you are calling and a single circuit
  is used, leaving other circuits available to
  other subscribers.
• LAN switches, also called switching hubs, look
  much like standard wiring hubs. The switching hub
  examines the data link header of the packet and
  obtains the destination address. The switch then
  establishes a dedicated connection between the
  senders port and the recipients port, and the
  two communicate.

39
A LAN Switch
Common Server
Common Server
Switching Hub 10/100 Mbps
To/From Other Hubs
Conventional Hub-10/100 Mbps
Conventional Hub-10/100 Mbps
To Workstations or Segment Servers
To Workstations or Segment Servers
40
Some Benefits of Switching Hubs
Efficient network segmentation to balance
traffic. Good price/performance. Availability of
shared and dedicated bandwidth. Support of new
technologies such as asynchronous transfer mode
and other network protocols. Preservation of
investment in cabling and LAN adapters. Devices
can communicate at NIC speeds as long as
necessary. Provide bridging and routing as well
as integration with faster technologies, such as
FDDI, 100 Mbps ethernet. Multiple connections can
be established concurrently. Ability to define
virtual LANs (VLANs)
41
Virtual LANs

• The benefit of VLANs is the ability to collect
  existing computers into a VLAN and to move
  workstations while maintaining their VLAN
  connection. Because VLANs are identified by
  addresses or switch ports and not by physical
  connections, LAN nodes can be moved from one
  location to another and remain on the same VLAN
  without having to change connections at the
  wiring hubs.
• A variety of methods are used to determine how
  users are grouped into a VLAN
• Port
• MAC Address
• Packet Tagging
• Network layer addresses

42
LAN-to-Host Connections

• The Host as a LAN Node
• Some hosts have the ability to connect to the LAN
  as a node. This is the most effective way to
  establish the connection.
• Asynchronous Connections
• Virtually every computer has the ability to send
  and receive by using an asynchronous data link
  protocol. Because most computers support the
  asynchronous data link protocol, it is sometimes
  used to link a microcomputer to a host. Usually,
  a microcomputer attached to a host asynchronously
  operates in one of two modes file transfer or
  terminal emulation.
• Dedicated Connection Per Microcomputer
• One way to connect a LAN node to a host is to
  provide a dedicated connection between a port on
  the host and each microcomputer needing a host
  connection. A dedicated connection provides
  direct host access, and the microcomputer does
  not use LAN resources for communicating with the
  host.

43
LAN-to-Host Connections (cont.)

• A dedicated connection has several disadvantages.
  First, as with all asynchronous connections, the
  sped of the link is slow. These connection speeds
  can be over 100,000 bps, but typically for
  microcomputer connections they are 33.6 Kbps or
  less. If many LAN nodes must communicate with the
  host, many host ports are required.
• Multiplexing
• A multiplexer is a hardware device that allows
  several devices to share one communication
  channel. Multiplexing is typically used to
  consolidate the message traffic between a
  computer and several remotely located terminals.
  This technique can also be used to allow several
  microcomputers to share a communication link to a
  host processor.
• Shared Asynchronous Connections
• In some applications, each LAN node needs
  occasional access to the host, but the number of
  concurrent connections is far fewer than the
  number of LAN nodes. In such situations, a
  dedicated line per node is excessive. A better
  solution is to share asynchronous connections.

44
Common Features of Asynchronous Communications
Software
Scripts Mouse support File transfers (CompuServe,
Xmodem, Ymodem, Kermit) Terminal emulation (ANSI,
DEC VT220, IBM 3101, TTY) Electronic mail Phone
directory Capture of data to a disk Text
editor Password security
45
Interconnection Utilities

• File Transfer Utilities
• File transfer utilities allow you to move files
  between network nodes. File transfer capabilities
  are an intrinsic part of many routers part of
  the TCP/IP protocol suite is a file transfer
  capability.
• Remote Logon
• A remote logon essentially establishes a remote
  user as a local user on the remote node. Once a
  user has successfully logged onto the remote
  node, commands issued by that user are processed
  and acted on by the remote node rather than by
  the local node.
• Remote Access
• Through remote connections, LAN administrators
  can resolve problems from home or other work
  locations users can perform some of their work
  at home and telecommute and travelers can
  conduct work while away form the office.

46
Interconnection Utilities (cont.)

• Integrated Services Digital Networks (ISDNs)
• One objective of ISDNs is to allow international
  data exchange. This requires interfaces between a
  number of national and regional providers of such
  services. The first mission of the ISDN program
  has been to define the functions and
  characteristics of the network and to establish
  implementation standards.
• ISDN was the first high-speed alternative to
  switched, analog connections for Internet access.
  ISDN when used for Internet access usually
  provides speeds of 128 Kbps.
• Digital Subscriber Lines (DSL)
• Digital subscriber lines are an emerging service
  that provides much faster transmission rates than
  analog modems and ISDN.

47
Assignment/Exercise 5/6 Local/Wide Area Networks

• Who is Robert M. Metcalfe? What is/are his
  contribution/s to the field of local area
  networks (LANs)?
• Read Metcalfe and Boggs paper entitled Ethernet
  Distributed Packet Switching for Local Computer
  Networks, Communications of the ACM, Vol. 19,
  pp. 395-404, July 1976 and comment on its
  relevance or irrelevance as far as local area
  networking is concerned in the 2000s. (Note
  Metcalfe and Boggs original Ethernet paper is
  available for free in some Internet sites use a
  search engine and search for the the article
  title - in quotation marks.)

48
Assignment/Exercise 5/6 Local/Wide Area Networks

• What is the Institute for Electrical and
  Electronic Engineers (IEEE) Committee 802? What
  are its functions and sub-committees?
• Describe the following (potential) wide area
  network physical/data link layer connection
  technologies
• plain old telephone system
• Leased line (64kbps, T1/E1, fractional T1/E1,
  etc.)
• Integrated Services Digital Network (ISDN)
• Digital Subscriber Lines (DSL)
• Frame Relay
• Asynchronous Transfer Mode (ATM)
• What is a repeater, a bridge, a switch, a router
  and a gateway? In what OSI layer does each of
  these devices functions?
• Submit a 2-4 page write-up of your findings (be
  brief but concise!)
• Due 19/26 Jan. 2002

49
Assignment/Exercise 5/6 Local/Wide Area Networks

• Familiarize with USENET Newsgroups related to
  local area networks, that is
• comp.dcom.lans.ethernet
• comp.dcom.lans.fddi
• comp.dcom.lans.novell
• comp.dcom.lans.token-ring
• comp.dcom.lans.misc
• etc
• Tip For starters, read the Frequently Asked
  Questions (FAQ) for each area.
• Where can you access USENET News? A lot of sites
  in the Internet
• Example http//www.google.com (then click on
  Groups)
• Submit (none)
• Due (N.A.)