University Campus, Hamilton

School of Computing

Building LANs

Topologies

Contents

1. NETWORKS...............................................................................................................................2

2. PHYSICAL STRUCTURES........................................................................................................2

3. CATEGORIES OF NETWORKS................................................................................................7

4. THE INTERNET..........................................................................................................................9

EXERCISE....................................................................................................................................11

5. NETWORK DEVICES...............................................................................................................12

EXERCISE....................................................................................................................................15

6. IMPLEMENTATION OF A BUS TOPOLOGY..........................................................................16

7. IMPLEMENTATION OF A RING TOPOLOGY.........................................................................21

8. IMPLEMENTATION OF A STAR TOPOLOGY........................................................................26

Building LANs Bell College

1. Networks

A network is a set of devices (often referred to as nodes) connected by communication

links. A node can be a computer, printer, or other device capable of sending/receiving
data generated by other nodes on the network.

Most networks use distributed processing, in which a task is divided amongst multiple
computers.

A network must be able to meet certain criteria such as performance, reliability, and
security.

2. Physical Structures

Type of Connection

A network is two or more devices connected through links, where a link is a

communications pathway that transfers data from one device to another. For
communication to occur, two devices must be connected in some way to the same link
at the same time. There are two possibilities for the type of connections: point-to-point
and multipoint.

Point-to-Point
A point-to-point connection provides a dedicated link between two devices. The entire
capacity of the link is reserved for transmission between those two devices. Most point-
to-point connections use an actual wire or cable to connect the two ends, but
alternatives such as satellite are possible.

Multipoint
A multipoint connection is one in which more than two specific devices share a single
link. The capacity of the channel is shared, either spatially or temporally. If several

Topologies page 2
Building LANs Bell College

devices can use the link simultaneously it is a spatially shared connection, whereas if
users take turns it is a timeshare connection.

Physical Topology

The term physical topology refers to the way a network is laid out physically. Two or
more devices connect to a link, and two or more links form a topology. The topology of a
network is the geometric relationship of all the links and linking devices (nodes) to one
another. There are four basic topologies: mesh, star, bus, and ring.

Mesh
In a mesh topology, every device has a dedicated point-to-point link to every other
device. The term dedicated means that the link carries traffic only between the two
devices it connects.

Topologies page 3
Building LANs Bell College

• Dedicated lines means that each connection carries its own data load. This
eliminates the traffic problems that occur when links must be shared by multiple
devices.

• A mesh topology is robust. If one link goes down, it does not result in the entire
network becoming unusable.

• Security and privacy is another advantage. Physical boundaries prevent other

users from gaining access to messages.

• Point-to-point lines make fault identification and fault isolation easier. Traffic can
be routed to avoid links with suspected problems.

The main disadvantage of a mesh is the amount of cabling and I/O ports required. A
mesh can therefore be prohibitively expensive and is usually implemented in a limited
fashion.

Star
In a star topology, each device has a dedicated point-to-point link only to a central
controller called a hub. The devices are not directly linked to one another. Unlike a mesh
topology, a star topology does not allow direct traffic between devices. The controller
acts as an exchange. If one device wants to send data to another, it sends the data to
the controller, which then relays the data to the other connected device.

Topologies page 4
Building LANs Bell College

A star topology is less expensive than a mesh. In a star, each device needs only one link
and one I/O port to connect it to any number of others. This makes it easier to install and
reconfigure.

Another advantage is robustness. If one link fails, only that link is affected and all other
links remain active. This also leads to easy fault identification and fault isolation. As long
as the hub is working, it can be used to monitor link problems and bypass defective links.

However, although a star requires less cable than a mesh, each node must be connected to
a central hub. More cabling is therefore required in a star than for other topologies such
as ring or bus.

Bus
The mesh and star topologies described point-to-point connections. A bus topology is
multipoint. One long cable acts as a backbone to link all the devices in a network.

Nodes are connected to the bus cable by drop lines and taps. A drop line is a
connection between the device and the main cable and a tap is a connector that either
splices into the main cable or punctures the sheathing of the main cable to create a
contact with the metallic core.

Topologies page 5
Building LANs Bell College

As a signal is transmitted along the backbone, some of its energy is transformed into
heat. It therefore becomes weaker and weaker as it has to travel further. For this reason
there is a limit to the number of taps a bus can support and on the distance between
those taps.

Advantages of a bus topology include ease of installation. Backbone cable can be laid
along the most efficient path, and then connected to the nodes by drop lines of various
lengths. A bus uses less cabling than a mesh or star topology.

A bus has numerous disadvantages.

• Disadvantages include difficult reconnection and fault isolation. A bus is usually

designed to be optimally efficient at installation and it can therefore be difficult to
add new signal devices.

• Signal reflection at the taps can cause degradation in quality. This degradation
can be controlled by limiting the number and spacing of devices connected to a
given length of cable. Adding new devices may require modification or
replacement of the backbone.

• A fault or break in the bus cable stops all transmission, even between devices on
the same side of the problem. The damaged area reflects signals back in the
direction of origin, creating noise in both directions.

Ring
In a ring topology, each device has a dedicated point-to-point connection only with the
two devices on either side. A signal is passed along the ring in one direction, from device
to device, until it reaches its destination. Each device on the ring incorporates a
repeater. When a device receives a signal intended for another device, its repeater
regenerates the bits and passes them along.

A ring is relatively easy to install and reconfigure. Each device is linked only to its
immediate neighbours and so to add or delete a device requires changing only two
connections. The only constraints are media and traffic considerations such as
maximum ring length and number of devices.

Topologies page 6
Building LANs Bell College

Additionally, fault isolation is simplified. Generally, traffic is circulating at all times and if
one device does not receive a signal within a specific time period, it can issue an alarm.
This alarm alerts the administrator of the network to the problem and its location.

Unidirectional traffic can, however, be a disadvantage. In a simple ring, a break in the

ring (e.g. a disabled station) can disrupt the entire network. Using a dual ring or a switch
capable of closing off the break can solve this weakness.

3. Categories of Networks

There are generally three categories of networks: local area networks, metropolitan area
networks, and wide area networks. Into which area a network falls is determined by its
size, its ownership, the distance it covers, and its physical architecture.

Local Area Network (LAN)

A LAN is usually privately owned and links the devices in a single office, building, or
campus. LANs are designed to allow resources to be shared between personal
computers or workstations. Resources include hardware (PC’s, Printers etc), software
(an application program), and data.

In many situations, a LAN links a workgroup of task-related computers. One of the

computers may be given a large capacity disk drive and may become a server to the
other clients. Software can be stored on the central server and used as needed by the
whole group. In such a situation the size of the LAN might be limited by licensing
restrictions on the number of users able to use the application software or to access the
operating system.

In addition to size, LANs are distinguished from other types of networks by their
transmission media and topology. Generally, a given LAN will only use one type of
transmission medium. The most common LAN topologies are bus, ring, and star.

LAN size is currently limited to a few kilometres. Typical speeds of LANs can reach
100Mbps.

Topologies page 7
Building LANs Bell College

Metropolitan Area Network (MAN)

A MAN is designed to extend over an entire city. It may be a single network, such as a
cable television network, or it may be a means of connecting a number of LANs into a
larger network so that resources can be shared LAN-to-LAN as well as device-to-device.
A company might use a MAN to connect the LANs in all its offices throughout a city.

Topologies page 8
Building LANs Bell College

Wide Area Network (WAN)

A WAN provides long-distance transmission of data, voice, image, and video information
over large geographic areas that may comprise a country, a continent, or even the entire
world.

In contrast to LANs, which depend on their own hardware for transmission, WANs may
utilise public, leased, or private communication equipment.

4. The Internet

A network is a group of communicating devices such as computers and printers. An

internet is two or more networks that can communicate with each other. The most
notable internet is the Internet itself, a collaboration of more than hundreds of thousands
of interconnected networks.

Topologies page 9
Building LANs Bell College

Topologies page 10
Building LANs Bell College

Exercise

1. Name the four basic network topologies and give an advantage for each type.

2. For n devices, what number of cable links are required for a mesh, bus, star, and
ring topology.

3. Describe factors that determine whether a communication system is a LAN,

MAN, or WAN.

4. What are the advantages of a multi-point connection over a point-to-point

connection?

5. Which topology requires a central controller or hub?

6. Which topology requires a multipoint connection.

7 In a network with 25 computers, which topology would require the most extensive
cabling?

8. A point-to-point/multi-point* connection provides a dedicated link between two

devices (*choose one).

9. In a point-to-point/multipoint* connection more than two devices can share a

single link (*choose one).

10. A break in a mesh/bus/star* topology stops all transmissions (*choose one).

11. For each of the following networks, discuss the consequences if a connection
fails:
a. Five devices arranged in a mesh topology.
b. Five devices arranged in a star topology (not including the hub).
c. Five devices in a bus topology.
d. Five devices in a ring topology.

12. Sketch a hybrid topology with a bus backbone connecting two ring backbones.
Each ring backbone connects three star backbones.

13. When a party makes a local telephone call to another party, is this a point-to-
point or multipoint connection? Explain your answer.

Topologies page 11
Building LANs Bell College

5. Network Devices

Networking devices include: repeaters, hubs, bridges, two- and three-layer switches, and
routers. Repeaters and hubs operate at layer1 of the OSI model and bridges and two-
layer switches operate in the first two layers. Routers and three-layer switches
operate in the first three layers of the OSI model. The following diagram shows the
layers in which each device operates:

Repeaters

Repeaters are devices that operate only at the physical layer.

Signals within a network can only travel a fixed distance before attenuation endangers
the integrity of the data. A repeater receives a signal and, before it becomes too weak or
corrupted, regenerates the original bit pattern. The repeater then sends the refreshed
signal and, in effect, can extend the physical length of a LAN:

A repeater does not actually connect two LANs but rather it connects two segments of
the same LAN. A repeater forwards every frame and has no filtering capability.

Topologies page 12
Building LANs Bell College

Hubs

Hubs operate at the physical layer.

A hub is actually a multi-port repeater and is usually used to create connections between
stations in a physical star topology. A hub therefore regenerates and passes all packets
to all ports. Hubs can be used to create multiple levels of hierarchy, and this hierarchical
use can overcome the 100m length limitation of 10Base-T:

Bridge

A bridge operates in both the physical and the data link layers. As a physical layer
device it regenerates the signal it receives. As a data link layer device, a bridge can
check the physical (MAC) address (source and destination) contained in a frame.

The difference between a bridge and a repeater is that a bridge has filtering capability. A
bridge can check the destination address of a frame and decide if the frame should be
forwarded or dropped. If the frame is to be forwarded, the decision must specify the port.
A bridge has a table that maps addresses to ports.

Consider the following example:

Topologies page 13
Building LANs Bell College

If a frame destined for station 712B1345642 arrives at port 1, the bridge consults its
table to find the departing port. According to its table, frames for 712B1345642 leave
through port 1 and so there is no need for forwarding. The frame is therefore dropped.
On the other hand, if a frame for 712B1345641 arrives at port 2, the departing port is
port 1 and the frame is forwarded. In the first scenario, LAN 2 remains free of traffic
whereas in the second scenario, both LANs have traffic.

It should also be noted that the bridge does not change the physical addresses
contained within the frame.

Switch

A (two-layer) switch is a bridge with many ports and performs at the physical and data-
link layers.

A switch improves network performance when used in place of a hub since it forwards
packets directly to destinations and cuts out contentions (collisions). Neither bridges nor
switches will stop broadcasts (broadcast storms).

For this course, we will consider switches as operating at the data-link layer, although
there are types of switches that operate at the network layer also.

Topologies page 14
Building LANs Bell College

Router

The router learns network addresses by sharing network knowledge with other routers. It
builds a routing table dynamically. Routers can pass data between networks but not
pass unwanted data. They can be used as firewalls for security.

A router has four components: input ports, output ports, the routing processor, and
the switching fabric.

Input Ports
An input port performs the physical and data link layer functions of the router. The bits
are constructed from the received signal. The packet is decapsulated from the frame.
Errors are detected and corrected. The packet is ready to be routed by the network
layer. In addition to a physical layer processor and a data link layer processor, the input
port has buffers (queues) to hold the packet before being directed to the switching fabric.

Output Ports
An output port performs the same functions as the input port, but in the reverse order.
First, the outgoing packets are queued, then the packet is encapsulated in a frame, and
the physical layer functions are applied to the frame to create the signal to be sent on
the line.

Routing Processor
The routing processor performs the functions of the network layer. The destination IP
address is used to find the address of the next hop and, at the same time, the output
port number from which the packet is sent out. For this activity, the routing processor
searches the routing table.

Switching Fabrics
These will not be discussed in detail here, but include such as crossbar switch,
banyan switch, and banyan-banyan switch.

Exercise
1. Describe the purpose of repeaters, hubs, bridges, switches and routers within a
network. State the network layer(s) at which each of these devices operate.

Topologies page 15
Building LANs Bell College

6. Implementation of a Bus Topology

• all workstations connect to the same cable segment

• commonly used for implementing Ethernet at 10mbps
• the cable is terminated at each end
• wiring is normally done point to point
• a faulty cable or workstation will take the entire LAN down
• two wire, generally implemented using coaxial cable during the 1980's

The bus cable carries the transmitted message along the cable. As the message arrives
at each workstation, the workstation computer checks the destination address contained
in the message to see if it matches it's own. If the address does not match, the
workstation does nothing more.

If the workstation address matches that contained in the message, the workstation
processes the message. The message is transmitted along the cable and is visible to all
computers connected to that cable.

There are THREE common wiring implementations for bus networks

• 10Base2 (thin-net, CheaperNet) 50-ohm cable using BNC T connectors, cards

provide transceiver
• 10Base5 (ThickNet) 50-ohm cable using 15-pin AUI D-type connectors and
external transceivers
• 10BaseT (UTP) UTP cable using RJ45 connectors and a wiring centre

Physical Implementation of a Bus Network

Topologies page 16
Building LANs Bell College

The above diagram shows a number of computers connected to a Bus cable, in this
case, implemented as Thin Ethernet. Each computer has a network card installed, which
directly attaches to the network bus cable via a T-Connector.

It is becoming common to use 10BaseT (UTP) for implementing Ethernet LANS. Each
workstation is wired in star fashion back to a concentrator wiring centre (hub). The hub is
a multi-port device supporting up to about 32 ports. One of these ports is connected to a
server, or the output of the hub can be connected to other hubs.

Ethernet 802.3: Carrier Sense Multiple Access with Collision Detection

(CSMA/CD)

This protocol is commonly used in bus (Ethernet) implementations.

Multiple access refers to the fact that in bus systems, each station has access to the
common cable.

Carrier sense refers to the fact that each station listens to see if no other station is
transmitting before sending data.

Collision detection refers to the principle of listening to see if other stations are
transmitting whilst we are transmitting.

In bus systems, all stations have access to the same cable medium. It is therefore possible
that a station may already be transmitting when another station wants to transmit. Rule 1
is that a station must listen to determine if another station is transmitting before initiating
a transmission. If the network is busy, then the station must back off and wait a random
interval before trying again.

Rule 2 is that a station which is transmitting must monitor the network to see if another
station has begun transmission. This is a collision, and if this occurs, both stations must
back off and retry after a random time interval. As it takes a finite time for signals to
travel down the cable, it is possible for more than one station to think that the network is
free and both grab it at the same time.

CSMA/CD models what happens in the real world. People involved in group
conversation tend to obey much the same behaviour.

Topologies page 17
Building LANs Bell College

Physical Bus Cable Limits

10Base2 Thin Ethernet Network Layout

Limitations

maximum number of trunk segments = 5

• maximum trunk segment length = 607 feet (185 meters)

• maximum network trunk cable = 3035 feet (925 meters)
• maximum number of stations on a trunk segment = 30
• minimum distance between T connectors = 1.5 feet (0.5 meters)

Rules

• each end of the trunk segment is terminated in 50-ohms

• one of the terminators is grounded
• connector splices are kept to a minimum

Cabling

• BNC-T type connectors

• RG58-AU 50-ohm cable, 0.2"

Topologies page 18
Building LANs Bell College

10Base5 Thick Ethernet Network Layout

Limitations

• maximum number of trunk segments = 5

• maximum trunk segment length = 1640 feet (500 meters)
• maximum network trunk cable = 8200 feet (2500 meters)
• maximum number of stations on a trunk segment = 100
• minimum distance between transceivers = 8 feet (2.5 meters)
• maximum transceiver cable length = 165 feet (50 meters)

Rules

• each end of the trunk segment is terminated in 50-ohm

• one of the terminators is grounded
• connector splices are kept to a minimum

Cabling

• Transceivers 802.3
• 50-ohm cable RG-11
• Male DIX connector

Topologies page 19
Building LANs Bell College

10BaseT UTP Network Layout

Limitations

• maximum segment length of 100 Meters

• Hub to Hub or repeater to repeater links limited to 100 Meters

Rules

• star topology
• 4 repeater/5 segment rule of 10Base5 is retained
• only two nodes per segment are allowed

Cabling

• RJ-45 Connectors
• Category 3 UTP minimum, preferably Category 5

Topologies page 20
Building LANs Bell College

Bus Network Topology Summary

Advantages Disadvantages
Easy to
Limits on cable length and Workstation numbers
implement
Low Cost Difficult to isolate network faults
A cable fault affects all workstations
As the number of workstations increase, the speed of the network
slows down

7. Implementation of a Ring Topology

• workstations connect to the ring

• faulty workstations can be bypassed
• more cabling required than bus
• the connectors used tend to cause a lot of problems
• commonly used to implement token ring at 4 and 16mbps
• four wire, generally STP or UTP

Physical Implementation of a Ring Network

Each workstation is connected back to a Multiple Access Unit (MAU), which supports up
to eight workstations. Additional MAU are cascaded to provide greater workstation
numbers.

Topologies page 21
Building LANs Bell College

Wiring is performed in a physical star fashion, with cables wired directly from each
workstation back to the MAU.

IEEE 802.5 Token Ring

This protocol is widely used in ring networks for controlling station access to the ring. A
short message (called a token) is circulated around the ring, being passed from station
to station (it originates from a controller or master station which inserts it onto the ring).

A station which wants to transmit waits for the token to arrive. When the token arrives,
the station changes it from a token to a connector message, and appends its message.
This new message is then placed on the outgoing side of the ring.

Each station passes on received tokens if they have nothing to transmit. They monitor
connector messages to see if the message is addressed to them. If connector messages
are addressed to them, they copy the message, modify it to signify its receipt, then send
it on around the ring. Connector messages which are not addressed to them are passed
directly on to the next station in the ring.

When the connector message travels full circle and arrives at the original sending
station, it checks the message to see if it's been received. It then discards the message
and replaces it with a token.

Topologies page 22
Building LANs Bell College

Topologies page 23
Building LANs Bell College

Either

Or

Topologies page 24
Building LANs Bell College

Physical Ring Cable Limits

Token Ring Network Layout

Limitations

• maximum number of workstations = 96

• maximum number of 8228 MAU's = 12
• maximum patch cable distance between an 8228 MAU and a station (not
including 8' adapter cable) = 150 feet (45 meters)
• maximum patch cable distance between two 8228's = 150 feet (45 meters)
• maximum patch cable connecting all 8228's = 400 feet (120 meters)

Rules

• stations are connected into the jacks of the 8228 units

• patch cables interconnect RO to RI for 8228 units.
• the last RO is connected to the first RI to form a ring.

Cable

• patch cables generally type 6 (26 awg) or 1 (22 awg)

• type 1 for lengths > 66 feet (20 meters)
• IBM 8310574 MIC connectors
• alternatively, UTP with RJ45 connectors

Ring Topology: Summary

Advantages Disadvantages
Cable failures affect limited users Costly Wiring
Equal access for all users Difficult Connections
Each workstation has full access
Expensive Adaptor Cards
speed to the ring
As workstation numbers increase
performance diminishes slightly

Topologies page 25
Building LANs Bell College

8. Implementation of a Star Topology

• all wiring is done from a central point (the server or hub)

• has the greatest cable lengths of any topology (and thus uses the most amount
of cable)
• generally STP or UTP, four wire

Star Topology: Summary

Advantages
Easy to add new workstations
Centralized control
Centralized network/hub monitoring
Disadvantages
Hub failure cripples all workstations
connected to that hub
Hubs are slightly more expensive than thin-Ethernet

Topologies page 26