NETWORKING AND CONNECTIVITY
A network allows computers to be attached together to allow sharing of resources or central file storage. While a
network is a multi
user system, it is important to understand the difference between a mainframe system
The technology of networking is extremely complex and beyond the scope of this document and course.
This unit seeks to address the basics of networking from the user and setup level rather than the technical.
Mainframe and Mini Systems
A mainframe system utilizes one or more computers, to which are attached units called terminals. Terminals are
usually comprised of a keyboard and monitor attached to a serial communication unit. The terminal may contain
memory or a small
processor dedicated to its own functions. Terminals are not computers and can only be used
when connected to the mainframe. A simple way of looking at it is to imagine connecting more than one monitor
and more than one keyboard to your computer and give
a monitor and keyboard to each person. You would still
have only one computer. This shared computer would slow down as the number of users was increased. While a
mainframe system can be part of a network, it is not a network.
A network is simply a numbe
r of computers
connected together to enable them to communicate with each other and to share hardware and software
TYPES OF NETWORKS
Peer to Peer
Networks are simply this type of network. No special equipment is needed except a means of
ecting them together. We will discuss the hardware requirements later. Both the Mac System and Windows
Systems support peer to peer networking.
Server to Client
Networks (LAN/WAN) bring this a step further. In a Server to Client system, one or more
puters are assigned a dedicated function and resources are centralized. This type of network has higher
security requirements and requires the Servers to run a Server Operating System. This type of system is referred
to as a Local Area Network or LAN. L
arger networks that cover larger geographic areas are usually called Wide
Area Networks or WAN. In a Server to Client network, servers are referred to as “Servers” and clients are the
individual computers called “Workstations”. These terms refer to the r
ole that this equipment plays within the
network rather than the type of hardware. LANs and WANs may contain Peer to Peer Networks, Main Frame
Systems and other LANs as clients.
Classification of Networks
Intranets and Internet
This unit extends the conc
epts of the previous unit to include connectivity with other systems. Let’s deal with
LAN (Local Area Network), that can be either peer
server, or a combination of both. A LAN is
usually limited to a single site.
WAN (Wide Area Ne
twork) is essentially the same as a LAN but covers multiple sites owned by the same
Intranet is a LAN or WAN that uses a standard Internet Browser and Email structure to make it appear as an
Internet connection and provides a seamless interface wit
h the Internet.
Internet is the connection of multiple LANs or WANs for the purpose of sharing resources.
World Wide Web (WWW) is usually used to describe the internet where multiple pathways (mesh) are
available for routing and connecting.
y computer can be networked as long as certain hardware and system software requirements are met.
Computers do not have to be of the same type. The table below shows you some methods of connecting
computers used today.
Direct Cable Connect
Null Modem Cable
Direct Cable Connect
Universal Serial Port Cable
Direct Cable Connect
Firewire Network Cable
Network Interface Card (NIC)
Network Card Driver
Twisted Pair (
Network Interface Card (NIC)
Network Card Driver
Network Interface Card (NIC)
Network Card Driver
Fiber Optic Cable
RF Interface Card
Radio Frequency Wireless
EHF Satellite Com.
As applied to actual network cable, media refers to the actual material and conductor format
within that cable. There
several in use today.
Similar to telephone cable, this has outer protective cover, within which are pairs of
insulated copper conductors. There are generally 4 pairs. The conductors of each pair
are twisted together in a specified number of twists per
inch. This controls interference
and insures proper bandwidth. Twisted pair cables are terminated using an RJ
modular connector system.
Similar to that used in television, coaxial cable contains an outer protective cover, with
which is a woven copper web forming a flexible tube (shield) extending the full length of
the cable. A central insulated conductor run
through the center of the cable. Coax is
terminated with BNC connectors.
This cable contains
an outer protective cover, within which is a bundle of glass or plastic
strands or light conductors. The cable is terminated using a complex process of polishing
strand ends and insertion into special connectors.
WARNING: In most locations, fire codes
cable running through air spaces
to be covered with
(Plenum Cable), since the plastic used on regular cable presents a fire and smoke hazard.
This is a term used to describe the actual way that wired network devices are
While there are many wiring schemes, only the more common ones will be discussed here. The topology used
can have an impact on the type of network system, the NICs, equipment and media to be used, as well as its
performance. For this reason,
choice of a topology must be made carefully. While smaller networks tend to use a
single topology, more complex networks are often designed around more than one topology in order to optimize
the advantages of each. While there are variations of each, I
will limit this presentation to the major pure
topologies. The table on the next page outlines the basic topologies used in designing networks. These are
sometimes used in co
mbinations, in order to exploit the advantages of each.
This is a simple passive system. Computer NIC cards
are connected using “T” connectors and coax cable. The
“T” of the first and last computer is terminated with a
small resistor called a “Terminator”.
This gets its name from the fact that each workstation’s
cable meets at a central point called a hub. The hub is a
device with a fixed number of ports that regenerates the
signal and prevents signal collisions.
A ring is like a bus with both ends connected together.
Signals are passed around the ring from one computer to
the other in one direction.
Token Ring NICs
Utilizes IR, RF, MW, Cellular or Satellite links to allow
mobility and connectivity with existing LANs or WANs.
and wireless NIC cards
Using multiple NICs to provide multiple pathways
Required equipment as
Combining multiple topologies to optimize each.
Required equipment for
each type as listed
At this point, we must make it clear that a number of computers just connected together do not constitut
unctional network. The physical connections discussed above present the simple part of networking. A
functional network requires a complex marriage of both hardware and software working together in order to setup
uniform communication. In the rest
of this unit, I will be addressing this complex issue. At times, I will resort to
oversimplification for the sake of understanding and brevity. The network technician must have an understanding
of the functions of all network components on both the hardwa
re and software level since they are intertwined.
Network Adapters or Network Interface Cards (NICs)
These may take several physical configurations:
Standard PCI board that inserts into the bus connector.
Onboard integrated type cons
isting of a chip on the motherboard.
PCMCIA Card type for notebooks
External USB unit.
Regardless of the physical configuration, NICs have a specified job
convert the signals from a form being used
by your computer to a uniform form used by the entire ne
. To make this possible, your computer must load
a software driver that enables your operating system to acknowledge the presence of the NIC and to
communicate with it. Like all devices, a unique address and an interrupt must be assigned to it. Lik
network protocol stack must be bound to the card's driver. The protocol (to be discussed later) sets up
communication rules or a common network language that will govern the shape of the information packet.
This is usually a resist
or set into a male BNC connector or on the board itself. It is used to terminate both ends of
a bus network backbone line. Without a terminator, the signal would run to the open end of the cable, then they
backup, usually out of phase with the oncoming s
ignal, thus causing collision cancellation. This is known as
Sometimes known as concentrators, hubs are used in star topologies to reconstruct, and amplify signals and
prevent collisions. Each hub forms a logical network entity called
a “node”. A network using a hub is called a
” because when a computer sends a packet, it goes to all the computers. All reject it except
he computer with the NIC of the proper address. This happens whenever a packet is sent. Hubs can
come in 2,
4, 8, 16 and 24 ports. Often a crossover port or BNC connector is also provided to allow other hubs (nodes) to be
added to form a tree structure.
A switch performs the functions of a hub but memorizes the port where it initially find
s the requested station so it
can send other packets directly and quickly there. Networks employing switches, rather than hubs are known as
”. Switches are available in three classes and when used with intelligent NICs offering
nal features like isolation of certain ports and password protection. Two networks may actually share a
switch in complete isolation from one another.
Routers and Gateways
Routers are kind of super switches, providing all of the functions of a top class
switch but use logical software
algorithms to compute the best routing pathways between computers. They can also figure out alternate routes.
Routers may also contain security and firewall protection (hacker protection). Routers are used where resources
must be shared across many networks to insure fast and efficient packet routing. Gateways are routers that
connect networks using different protocols.
While not part of the network, these
Uninteruptable Power Units
Auxiliary Power Units
ure a constant,
clean source of power to crucial network components. They generally contain a battery charger and a set of
batteries that remain charged. In the event of a power failure, an inverter circuit converts the DC Battery power to
120 VAC power
line power to provide a limited supply of power. Some units will automatically shutdown
computers or make a modem call. Larger networks use these in conjunction with power alternators to insure
continuous power. Servers, switches and critical workstatio
ns should have these units.
These are simply computers connected to a server to client network.
Ready Devices contain their own NICs and server firmware to connect directly to the network system.
Dedicated Network Servers
rvers must contain
enough storage to hold the network operating system as well as the user security
and network structure database. It also holds programs and data made available to the network clients.
While the file server in a small network often carr
ies out many of the following functions, larger networks
require one or more of the following dedicated servers. A Server
Client network must have at least one file
server at the primary node.
Print Servers manage print jobs sent to network printers th
at are not network
Backup Servers contain stacks of storage like Hard Drives, CD
R Drives, or Tape Drives with appropriate
software and firmware to allow network
wide backup of files. Some have automated tape
ROM Servers (juke
box) provide stacks of CD
and server firmware to allow plugging right
into the network.
Data Servers contains large storage areas for database applications where the file server storage would not
Host Configuration Protocol (DHCP)
connect to the Internet line and lease out
IP addresses to workstations, allowing them to be serviced by a single connection.
mail Servers manage internal email accounts across a network.
Proxy Servers support firewall
software and manage Internet requests.
Peer Servers are workstations that are sharing resources across the network. In such a case, the station
functions as the server for those resources and must be active to make these resources available.
Web Servers a
re servers that host a web sites and associated services.
Creating a Dedicated File Server
Select a Network Server Operating System based on the future size of your network, the number of users and
the required security level. While a larg
er secure LAN may dictate Novell Netware or Windows NT Server or
Windows 2000 Server Edition, very small networks may be satisfied with Windows 98se, me or Windows
Select a computer suitable for a file server based on the processor, mem
ory and storage requirements of your
server operating system and the number of users to be accommodated.
Partition and format the Hard Drive(s) for your server operating system according to instructions. You will
have to specify network volume names that
will identify these drive partitions for the workstations.
Install the NIC Card or NIC Cards, as needed and required drivers.
Install your network server operating system. Depending on your system, the process and sequence will
vary. You will give your s
erver a name, define its volumes, protocol and bind to the card drivers (see
Setting up Workstation.
) You will then have to define domains, workgroups, etc.
You will then define print queues, print servers and network printers. Other auxil
iary servers are handled
much the same.
Next define templates, scripts, restrictions, groups and rights.
Finally, create your users and user passwords and assign them to groups, from which they will inherit rights.
You can grant additional rights, where n
eeded, to particular users.
Test the network and do a network backup.
all network workstations must be configured using the following steps.
Install the appropriate Network Interface Card (NIC) on the computer to be used as
Install the driver for the card.
Install the network protocols required to communicate over the network. Some protocols require a
unique IP address for each station.
Install the client software to communicate with each type of server.
there are variations depending on the hardware platform and operating system, generally the procedure for
making a computer a network workstation is remarkably similar. The following steps will talk you through the
process. I am using an Ethernet twiste
d pair Star LAN topology here, since it is one of the most common
Install the NIC
. Turn off your computer’s power switch and open the case, if necessary. If you are installing a PCI
NIC, locate an available PCI slot (white connector) and remo
ve the slot tab. Retain the screw. Insert the card
using even pressure making sure the card’s metal tab occupies the space once occupied by the tab that you
removed. When the card is fully inserted replace the screw. Some computers use clips instead of
one end of your network patch cord into the RJ
45 connector on the card’s tab. The other end is inserted into an
available port in the hub or RJ
45 Jack Box. A green light should appear on the card’s tab indicating electrical
Each NIC has a unique address that is attached to a packet and identifies your station. It’s your
stations “Return Address”
. Install the Driver. Your card manufacturer will supply a driver on disk or CD. Follow
the directions. (A plug ‘n Play operati
ng system will talk you through much of the following procedure and assign
the Address and IRQ, but you will need to select options.) In Windows, you can check this installation in the
Install a Protocol
. A protocol is the language used
across the network. Like any language, a protocol has a
grammar and syntax or a set of rules governing the shape and handling of packets of information.
defines the order in which this binary information is interpreted from the packet or st
ring of “0”s and “1”s. A
network must use a common protocol in order to communicate. A workstation communicating with more than one
network with different protocols must declare a separate protocol for each. During protocol declaration, a process
“binding” will take place. Binding attaches the protocol stack and instructions to your NIC Card’s driver.
This is what tells your card how to shape the outgoing packets and interpret incoming packets for that protocol.
“Protocol” is a complex topic th
at cannot be treated full here. Full treatment of the topic is the domain of a
network engineer’s course. Suffice it to say that
Protocols differ widely in efficiency, routing capability and error
While most protocols can allow packet
s to route through hubs and switches
protocol must be used to pass packets through routers and gateways to networks of different types
popular protocols are listed below.
TCP/IP is recommended due to the fact that it is a routable
protocol that will
allow Internet connections.
Install the Client Type
. In order for a station to see a server and be recognized by that server, you must declare
a client type. The client type is
always based on the type of server operating system
u may declare more
than one type of client. For example, if your network has a Windows NT Server and a Netware Server and you
wish to use both, you would declare two client types:
Client for Microsoft Networks
to use the NT Server and
Microsoft Client fo
to see the Novell Netware Server.
Setting Up a Peer to Peer Network
A peer to peer network is a network without a dedicated server. Instead of centrally locating resources and
services, they are spread among the workstations. In a sense, each
workstation can act as a server to the others
when they request resources. Setting up workstations for this type of network is the same as setting up any
workstation. However, there are additional steps that must be performed: On the host computer (the
1. Sharing must be turned on in the network properties.
2. The device to be shared must be given a recognizable name and made sharable for read
only or read
and given password protection. Only users having this passwor
d will be able to use the resource. The client
computer must also be set up. Let’s look at the following example. Computer A will share its HP Laser Printer
(HP) and its CD
ROM (D:). Computer B will share its entire Hard Drive (C:). On Computer B you m
Computer A’s Printer as a network printer known as HP and Computer A’s CD
ROM mapped as E: since
Computer B already has a CD
ROM of its own (D:). On Computer A, Computer B’s Hard Drive is mapped as E:
since it has a CD
ROM called D:. While a
workstation can be dedicated to act as a server, the operating system
does not have the user level security that a server system has. Peer to Peer Networks may also be clients in a
Due to the increased cost, lowe
r transmission speeds and lower reliability of wireless connections, wireless
networks are almost always part of a larger hard
wired network. Wireless network connections afford a freedom
that wired connections do not under the following conditions.
mobility is desired or necessary
Where connections are only temporary.
Where skilled labor for wiring is not available
Where the building structure does not lend itself to wiring or wiring would be too labor intensive.
Radio Frequency (RF)
RF Transceiver NICs
RF Transceiver called
. These act as wireless hubs and
form a junction point between a wired and wireless network.
Microwave Link (MWL)
Point to point dish
Satellite Link (Sat
Satellite Dish Antenna and uplink
Although we can connect any type of hardware using any operating system and any netw
ork type, in order for an
Intranet, the Internet or the World Wide Web to work, the following rules must be understood.
A common routable protocol must be used. Remember that the protocol is a sort of language used across a
network. This language has a c
ommon set of rules governing the stream of contents, shape and order that
information must be in to be interpreted by equipment using the network. Part of this information is routing
information that functions like the postmark, address and return address
on a postal envelope.
While graphic images and other file forms may be ported across the Internet, the system is a mainly a UNIX
based, tagged text
oriented system. The system relies on a program called a browser to translate the tagged
text into the visu
al graphics you see as a web page. The browser interprets the tagged text into commands
causing the browser to reconstruct the original web page in the client’s workstation. The most common
tagged text language in use today is called
nguage (html). The most common browsers
in use today are Microsoft
Connecting to the Internet and World Wide Web means connecting to a provider’s server. This web server
may belong to a single site or may act as a gateway
to a larger system. When this happens, the server’s
operator is known as an Internet provider. In reality, a typical session means connecting and bouncing from
server to server and network to network. Large providers like
America On Line
ntain a system
of servers and networks that offer a number of services in addition to web connection. These may include
email, chat services, phone services, search engines and travel services. Other providers like Hotmail or
Yahoo offer email and more l
A physical wired or wireless connection with the provider’s server must be established and maintained
throughout the session in order to provide throughput of information. This may be done using a temporary
up connection via phone li
ne or cellular phone network or a dedicated continuous connection. Some
Popular Methods of Physical Connection
Enable a connection over ordinary analog voice telephone
Speed: <56 Kbps.
device called a Modem
that uses audible tones to transmit
mation to the destination
modem that interprets it back into
Integrated Services Digital Network
ISDN lines are digital lines designed for higher bandwidth.
s: 64/128 Kbps.
384 and 1.544 Mbps
Employs a special ISDN modem
connected to digital dedicated
Asymmetric Digital Subscriber Line
Makes use of existing phone lines and available service from
Speed: 6 Mbps down, 640 Mbps
Employs a special modem and
ADSL, HDSL, VDSL, UDSL, SDSL
that vary in speed depending on
TV Cable Service
Uses TV Cable Distribution System
Employs a special Cable Modem
Uses small satellite dish antenna and associated equipment
Employs a special modem,
descrambler equipment and
Digital Leased Copper or Fiber Optic Line
44 Mbps. 24 channels
T2 Carrier, Digital Leased Fiber Optic Line
Speed: 6.312Mbps. 96 channels
T3 Carrier, Digital Leased Fiber Optic Line, Speed: 44.736
Mbps. 672 channels
Digital Leased Fiber Optic Line
Speed: 274.176 Mbps. 4032 channels
Requires CSU/DSU and DNS
Server as well as a LAN distribution
When using dedicated digital service lines, services require several providers beyond the provider of the Internet
and Email services. The local Phone Company provides the line from y
our location to the central office or point of
presence (POP) where the actual service provider sets op a connection. This is different from using a modem,
where you are using the public switched phone lines.
Internet Information Routing
mation along the is the process of getting information from its origin to its intended destination, much
the same as a letter going through the postal system. Just as the post office requires a standard amount of
certain information like the name and addr
ess of the sender and recipient, postmarks and barcodes, the electronic
packet must contain similar information. Unlike first class mail however, the actual message content, due to the
limited packet size, need not travel in the same envelope. As a matte
r of fact, during a given transmission,
different parts of the message content may actually be routed in different paths to its intended destination, where
it is reassembled. This is why, sometimes, web pages may appear to come in and fill the screen in a
While a full treatment of the process is rather detailed and complicated and required an understanding of the
various protocol layers, a glimpse of this process will be discussed here. To provide for this process, a un
addressing system must be employed. This is accomplished by providing each node (addressable device) with
what is called a unique IP address. The rules for this process are part of the TCP/IP Network Protocol. No two
nodes may have the same addr
ess. You may have notices numbers like: 126.96.36.199 that is an example of
an IP Address. Addresses are assigned using a uniform system that designates different groups of addresses to
different types of equipment and networks. In order to create a
more recognizable addressing system, an IP
address is associated with a domain naming system. A DNS Server maintains a database for this purpose. This
is why we can use an address like:
rather than the actual IP address.
Security and Ethical Issues
Dialup Connections vs. Dedicated Digital Lines
(DSL, Cable, T1)
One must keep in mind that the longer you are connected to a network of any kind, the greater the possibility that
viruses may invade your computer. Therefore, while not exempt from the problem, dial up presents
the least danger. With the proliferation of DSL, Cable Internet, Satellite Internet and T Carrier Lines, the problem
of these invasions has gone up geometri
cally. Personal Firewall and Virus Prevention software is readily
available for a reasonable price and is a must for users having these continuous services.
While protection is necessary, like door locks, it would be a better world if it were not needed.
In the United
States, hacking and computer crime is a Federal Offense. This does not deter some unscrupulous characters
from taking liberties with other people’s sites and systems. Hackers use many explanations and excuses;
however, the fact remains th
at, if you enter any space uninvited, you are guilty of invasion.
A virtual space is no
different than a physical space and you don’t belong there.
Just imagine if you found a stranger in your home
when you go home today. Would you really accept the exc
use that, well, your lock was easy to pick? The
creation and propagation of viruses is property destruction. This electronic graffiti can be extremely destructive
and can cause irreversible damage that costs time and money and, in the long run, makes th
ings more expensive
for all of us.
Copyrighted software is another area of consideration. Does it hurt anyone if you make illegal copies of software?
After all, large software companies make a lot of money. Think about it. How do companies make up for
sales? They raise the price of software for everyone or they downsize by eliminating workers. Those workers
have families to feed.
Ethical behavior is the mainstay of a civilized society; we must all work to preserve it.