Experiment # 1

Introduction to Networks Topologies & Components

2-1 : Introduction
A computer network can be small and consist of only two computers, or it can be a
worldwide network such as the Internet. Due to today's high-speed electronics,
communicating with a computer on the other side of the planet can be virtually as fast as
communicating with a computer just down the street. Now let's take a look at the smaller,
interoffice-type networks called LANs (Local Area Networks).

2-2 : Objectives
At the end of this experiment, you will be able to:
• Describe what a Local Area Network is.
• Describe the different types of LANs.
• Describe some of the key components of a network system.
• State some of the benefits LAN technology offers to system users.

2-3 : Local Area Networks

A Local Area Network (LAN) is the interconnection for computers and not the devices
which are connected to it. Practically, however, a LAN is the interconnected computers and
their peripherals in an area that covers a small geographic area. Every node (a node is a
computing device or station that connects to the network at a single location) on the network
can communicate with every other node. Though the network may contain a central processor,
a central processor is not required. LANs can provide electronic exchange of various data
types including voice, computer data, word processing, facsimile, video, telemetry, and other
forms of electronic data.

LANs are geographically local, but a LAN may function in a fairly large area (such as a
university campus or among government buildings). The important thing to remember is that
any single point in the network can communicate with any other point, allowing independent
devices to communicate with each other.

8
2-4 : LAN Terminology

Node : A single device, such as a personal computer, connected to a network at a single

location. Nodes are also called stations or workstations.

Packets : Packets are data units, pieces of data, that are acted upon by the network. They pass
through various layers in the network architecture and are used to send information through
the network.

Personal computer (PC) : Another name for a microcomputer. The term PC has come to
mean a machine running the DOS operating system, but PCs are not limited to that operating
system. A PC is a computer that is self-contained, but limited in storage space and speed. It is
typically used by one person at a workplace.

Protocol : A protocol is the predefined standard or set of rules by which a function or service
is performed.

Server : a personal computer or specialized device that provides and manages access to
shared network resources, such as hard disks and printers.

Terminal : A device that allows input and output of data to a computer. The term is most
frequently used to describe a device that has a keyboard for data entry and either a printer or a
monitor for displaying data.

Workstation : IBM® computers or compatibles, Apple Macintosh® computers, or UNIX

machines from Sun Microsystems® and other vendors. Workstations actually compute and
execute applications; they are not passive parties to the computing process.

2-5 : LAN Topologies

LANs can be classified according to their structure. A network, in its simplest form,
consists of two computing devices that are electronically connected and are able to exchange

9
information. Communication between the central computer and the client machine is
accomplished by a direct line, RF (radio frequency), infrared, or through a modem.
The users may never know which type of network they are connected to. Often, the only
knowledge the user has of the system is the interface they use (the screen they are looking at).
The following sections describe different LAN types: single node, star, ring, bus, and
hierarchical.

2-5-1 : SINGLE NODE

Figure 2-1 shows a central computer (or server) and terminal. (NOTE: The terms
terminal and workstation are often used interchangeably.)
This server, or host, is the main repository of information for the network. The terminal,
sometimes called a "dumb terminal," does not usually contain the files and other information
needed to function independently

Figure 2-1 : A single-node network

A terminal can access files and applications from the host, work on them, and return them to
the host. However, some terminals do contain their own applications and files. In this case,
they may use the central computer only for backing up files on its hard drive or using its
printer or other peripheral device.
A dumb terminal is the cheapest computer device to use on a network since you don't have to
pay for the processing power of a stand-alone computer, but a dumb terminal relies on the
resources of the central computer.
The central computer (host) need not be a large computer. In fact, it may be a very small
computer (microcomputer) such as the one you would find on a desk or in someone's home.
In many cases, however, the central computer is larger (mini computer to mainframe), and
capable of hosting hundreds of terminals.

10
2-5-2 : STAR ARCHITECTURE
The star architecture, shown in Figure 2-2, has a central hub to which all of the nodes
are attached, usually by copper cable or fiber optics.
The hub is the focal point of the star configuration, and all network traffic must pass through
the hub. In this configuration, multiple nodes are attached to the central computer, rather than
just two nodes, as shown in Figure 2-2. The star network represents a cost-effective way to
share information among many users. Typical uses include airline reservation counters, or
small business offices in which several people must have access to common applications and
files.
These days, many or all of the nodes on the network have their own intelligence, though
they can be dumb terminals with all of the power kept in the central computer. If the nodes
have their own intelligence, there is little need for a central computer and the intelligence can
be spread out across the nodes.

Figure 2-2 : A star network

Connections are accomplished as in the single node example that uses direct lines or modem
connections. The way in which they are connected may also vary. Each node is usually
connected to the central computer.

2-5-3 : RING STRUCTURE

In a ring topology, each station is electrically, or optically, connected to (two neighbors,
one "upstream," and the other "downstream," in a ring as shown in Figure 2-3. A token, or

11
small message, is continually passed from node to node around the ring. Although the token
passes through every node on the network, it can travel the entire ring very quickly because
very little processing of the token occurs at each node. Because of the use of the token,
networks with ring topologies are often called Token-Rings.
To transmit on a Token-Ring network, a node waits to receive the token without any message
attached to it. The node then appends its message to the token which makes its way around
the ring as just described.

Figure 2-3 : A ring network

The destination station receives the message, that is, copies it from the network into its
memory, and changes the token to indicate the message has been received. The changed token
then continues around the ring back to the sending node. When it receives the changed token,
the sending node "knows" the message has been received. It may even read back the message
it sent to confirm it traveled the ring intact.
The token-passing scheme is "fair" in that it insures that each node on the network is allowed
to send a message in turn, before any node is allowed to send again. Token-passing schemes
can also support the notion of priority, where all messages of higher priority are sent before
messages with lower priority. Although the nodes on a Token-Ring are connected electrically,
or optically, in ring structure, physically, the nodes are almost always connected in a star to a
hub or MAU (multi station access unit) as shown in Figure 2-6.
The MAU connects nodes to the ring as they are activated or turned on, but makes certain any
node which malfunctions is removed from the ring, thus preventing a malfunctioning node
from bringing down the network. A malfunctioning node can interrupt connectivity, but by
using redundant connections or rerouting the signal traffic, the network's connectivity can be
maintained. IBM established a standard for its Token-Ring network which has since become

12
rather popular. IBM's Token-Ring can operate at either 4 Mbps (million bits per second) or 16
Mbps.

2-5-4 : BUS STRUCTURE

The bus structure, shown in Figure 2-4, consists of two ends usually connected by some
sort of cable. Nodes tap into a cable (called the backbone). Signals travel along the backbone
after they are generated and the nodes "listen in." The nodes recognize signals that carry their
particular address designation. For instance, a node may be designated as machines.
If a signal is traveling along the backbone, all of the nodes sense the signal, but only the node
with the correct designator actually receives the signal. It is possible to make the signal
accessible to all nodes on the backbone if all of the machines receive the signal. For instance,
all (or some) departments of a company could see an important memo.
A good example of a bus structure network is a typical Ethernet network in which nodes tap
into the backbone. This type of network is only limited by the number of nodes that can be
added to the backbone, and by the length of the backbone (though the length can be increased
and augmented by certain means).
Intelligence can be spread across a bus network so that even when one or more nodes are out
of action, the overall network can continue to function.

2-5-5 : HIERARCHICAL STRUCTURE

Hierarchical network systems, as shown in Figure 2-5, use the capabilities of all nodes
on the network to their fullest extent. This structure distributes the intelligence across the
network, and provides connectivity between all of the nodes in the system.
Typically, at the top of the network are large, high-speed computers that are used for
storage and for processing complicated functions. The next level consists of mid-range
machines that send data to the large computers. Feeding into the mid-range computers are
smaller computers (minis and micros). The smaller machines may be dumb terminals, but
increasingly, they are sophisticated individual workstations that perform with the capacity and
speed of the mainframe computers in use not so many years ago.

13
 Figure 2-4 : A bus structure network

Figure 2-5 : A hierarchical network structure

2-5-6 : MESH TOPOLOGY

The mesh topology is one where there are at least two paths to and from ever node. See
Figure 2-6. In some mesh topologies every node is connected to all the other nodes. The mesh
topology is useful for hostile environments where connections are easily broken. This way, if
a connection is broken there is always another path so the network can continue working. If
the mesh topology has every node connected directly to every other node, then obviously the
network must be small or the wiring will quickly become a horrible mess.

Figure 2-6 : The Mesh topology

2-6 : Network Size
Networks can also be classified by size, as follows:
LAN : Local Area Network usually covers less than 1 km.

14
CAN : Campus Area Network covers an area of 5 to 100 km and are privately owned.
MAN : Metropolitan Area Networks are from 2 to 100 km.
WAN : Wide Area Networks span more than 100 km and can encompass the entire world.

2-7 : Interface Requirements

Multiple node networks require that each node (computer) have. a network interface adaptor,
often in the form of a card (NIC), installed in it. This circuitry handles the actual sending and receiving
of data between the communicating computers.
The network interface card performs the following functions:
• Physically attaches the computer to the network.
• Frames the data for transmission as digital signals.
• Decodes received signals and converts them back into data.
• Performs error detection and possibly retransmission.
The interface card must match the type of media that the network will use. The media will
move the data between the computers.

Example: If you will be using fiber optic cable, then the NIC must have fiber optic
connectors on it. Figure 2-7 shows a NIC with an RJ45 connector and a BNC connector, but
many Ethernet cards have only one connector or up to three connectors to choose from.

Figure 2-7 : Network Interface Card (NIC)

15