Network topology architecture

Network Topology andNetwork Topology and
DesignDesign

ObjectivesObjectives
• Discuss the different physical topologies
• Determine which type of network media to use
given a set of requirements
• Consider performance requirements and
improvements for given situations

Network TopologyNetwork Topology
• Topology
There are two types of topology:
physical and logical.
• The physical topology of a network refers to the
configuration of cables, computers, and other
peripherals.
• Logical topology is the method used to pass the
information between workstations.

Physical Topologies:Physical Topologies:
BusBus
• All devices are connected to a central cable,
called the bus or backbone. Bus networks are
relatively inexpensive and easy to install for
small networks. It has a single cable with
terminators at each end.

BusBus
• A bus topology connects all stations in a linear fashion
Figure 4-1: Bus topology
Terminator - A device that provides electrical resistance at the end of a transmission line. Its function is to
absorb signals on the line, thereby keeping them from bouncing back and being received again by the network.

BusBus
• Bus topology advantages:
– It is inexpensive
– It is easy to design and implement because the
stations are simply daisy-chained together
• Bus topology disadvantages:
– It is difficult to troubleshoot
– It requires termination

StarStar
• The star network configuration is the most popular
physical topology
• In a star configuration, all computers or stations are
wired directly to a central location:
– Concentrator (a.k.a. hub)
– Multistation Access Unit (MAU)
• A data signal from any station goes directly to this
central device, which transmits the signal according
to the established network access method for the
type of network
• The protocols used with star configurations
are usually Ethernet or LocalTalk

StarStar
Figure 4-2:
Star topology

StarStar
• Star topology advantages:
– A break in one cable does not affect all other
stations as it does in bus technologies
– Problems are easier to locate because symptoms
often point to one station
– The second-easiest topology to design and install
– Does not require manual termination
• Instead the media is terminated in the station at the
transceiver on the NIC and in the hub or MAU

StarStar
• Star topology disadvantages:
– Hubs, which are required for a star topology, are
more expensive than bus connectors
– A failure at the hub can affect the entire
configuration and all connected stations
– Uses more cable than bus topologies

Star/bus/TreeStar/bus/Tree
• Bus and star topologies can be combined to form
a star/bus or bus/star physical topology
• Hubs that have connectors for coaxial cable as
well as for twisted-pair wiring are used to form
these types of networks
• When different physical topologies are applied to
a network, the result is often called a mixed
media network

Star/Bus/TreeStar/Bus/Tree

Advantages of a Tree Topology
• Point-to-point wiring for individual segments.
• Supported by several hardware and software
vendors.

Disadvantages of a Tree Topology
• Overall length of each segment is limited by the
type of cabling used.
• If the backbone line breaks, the entire segment
goes down.
• More difficult to configure and wire than other
topologies.

RingRing
• A ring network is a network topology in which each
node connects to exactly two other nodes, forming a
circular pathway for signals - a ring. Data travels
from node to node, with each node handling every
packet.
• Because a ring topology provides only one pathway
between any two nodes, ring networks may be
disrupted by the failure of a single link.
•

RingRing
• A system of which each node or station is
connected to two others, ultimately forming a
loop (circular pathway for signals).
• Data are passed in one direction only, being
received by each node and then transferred to the
next node.
•

RingRing
• Data travels from node to node, with each node handling every packet.

RingRing
• Physical rings
– Most often seen in Fiber Distributed Data
Interface (FDDI) networks
• FDDI is a WAN technology
– Stations on a ring are wired to one another in a
circle around the entire network

RingRing
• Ring topology advantages:
– It prevents network collisions because of the
media access method or architecture required
– Each station functions as a repeater, so the
topology does not require additional network
hardware, such as hubs

RingRing
• Ring topology disadvantages:
– As in a bus network, a failure at one point can
bring down the network
– Because all stations are wired together, to add a
station the network must be shut down
temporarily
– Maintenance on a ring is more difficult than on a
star topology because an adjustment or
reconfiguration affects the entire ring

Considerations When Choosing a Topology:
• Money. A linear bus network may be the least
expensive way to install a network; you do not have
to purchase concentrators.
• Length of cable needed. The linear bus network uses
shorter lengths of cable.
• Future growth. With a star topology, expanding a
network is easily done by adding another
concentrator.
• Cable type. The most common cable is unshielded
twisted pair, which is most often used with star
topologies.

WAN TopologiesWAN Topologies
Figure 5-7:
WAN
topologies

Influence of the 5-4-3 Rule onInfluence of the 5-4-3 Rule on
TopologiesTopologies
• 5-4-3 rule states that between stations on a LAN, there can be no more
than five network segments connected, maximum number of repeaters is
four, and maximum number of segments with stations on them is three
Figure 4-3:
5-4-3 rule

Influence of the 5-4-3 Rule onInfluence of the 5-4-3 Rule on
TopologiesTopologies
Figure 4-4:
Mixed
topologies

Twisted-Pair CablingTwisted-Pair Cabling
• Common traits of all twisted-pair cabling
types and categories:
– The wires are copper
– The wires come in pairs
– The pairs of wires are twisted around each other
– The pairs of wires are usually enclosed in a cable
sheath individually and as a group of wires

Twisted-Pair CablingTwisted-Pair Cabling
• Crosstalk
– Signal bleed from one cable to another
– Usually occurs in poorly wired media
• Cancellation
– Insulates the signal from the effects of signal
bleeding

Unshielded Twisted-Pair (UTP)Unshielded Twisted-Pair (UTP)
• Cabling used for a variety of electronic communications
Table 4-1: Categories of
UTP

• UTP advantages:
– Thin flexible cable that is easy to string between
walls
– Most modern buildings come with CAT 5 UTP
already wired into the wall outlets or at least run
between the floors
– Because UTP is small, it does not quickly fill up
wiring ducts
– Costs less per foot than other type of LAN cable

• UTP disadvantages:
– More susceptible to interference than most other
types of cabling
• Pair twisting does help, but it does not make the cable
impervious to electrical noise
– Its unrepeated length limit is 100 meters

RJ-45 ConnectorsRJ-45 Connectors
• Registered Jacks (RJ)
– Type of telecommunication connector used for
twisted-pair cabling
– Typically RJ-45 connectors resemble the typical RJ-
11 connectors that connect the phone to the wall
• Difference between RJ-45 connectors and RJ-11 connectors is
that the former has eight wires (four-pair) and the latter four
(two-pair)
– Some RJ-11 connectors are used with three-pair (six-
wire) UTP

Shielded Twisted-Pair (STP)Shielded Twisted-Pair (STP)
• Cabling often seen in Token Ring networks
• Similar to UTP in that the wire pairs are
twisted around each other inside the cable
• The advantage of STP over UTP is that it has
greater protection from interference and
crosstalk due to the shielding

Shielded Twisted-Pair (STP)Shielded Twisted-Pair (STP)
• STP disadvantages as compared to UTP
include:
– A higher cost per foot
– The shield must be grounded at one end
• Improper grounding can cause serious interference
– Heavier and less flexible
– Because of its thickness, STP may not fit down
narrow cable ducts

Coaxial CablingCoaxial Cabling
• Consists of either:
– A solid inner core (often made of copper)
– Wire strand conductor surrounded by insulation
• The two most commonly used coaxial cable:
– Thicknet
– Thinnet

• Advantages of coaxial cabling on a LAN
include:
– The segment lengths are longer than UTP or STP
– Coaxial cable has greater interference immunity
than UTP
– Hubs between stations are not required

• Disadvantages of coaxial cable:
– Not as easy to install as UTP
– More expensive than UTP
– Supports a maximum bandwidth of only 10 Mbps
– Requires more room in wiring ducts than UTP
– Is relatively difficult to troubleshoot thinnet and
thicknet networks
– Connectors can be expensive.
– It is easily damaged and sometimes difficult to work
with, especially in the case of thick coaxial.
– Baseband coaxial cannot carry integrated voice, data,
and video signals.

Table 4-2: Coaxial cable types

Thinnet and Thicknet ConnectorsThinnet and Thicknet Connectors
• The most common connectors for RG-58 cabling
on thinnet networks are:
– Barrel connectors
– T-connectors
– Terminators
• BNC
– Hardware connector for coaxial cable with a
cylindrical shell with two small knobs allowing it to
be locked into place when twisted

Thinnet and Thicknet ConnectorsThinnet and Thicknet Connectors
• Attachment unit
interface (AUI)
port
– A 15-pin physical
connector
interface between
a computer’s
network NIC and
an Ethernet
networking that
uses 10Base5
coaxial cableFigure 4-6: Thinnet connectors

Fiber-Optic CableFiber-Optic Cable
• Carries light pulses rather than electrical
signals long its fibers
• Made of glass or plastic fibers, rather than
copper wire like most other network cabling
• Core of the cable is usually pure glass
– Surrounding the glass is a layer of cladding made
of glass or plastic, which traps the light in the core

• Fiber-optic cabling advantages:
– Can transmit over long distances
– Not susceptible to electromagnetic interference or
crosstalk
– Supports extremely high transmission rates
– Cable has a smaller diameter and can be used in
narrow wiring ducts
– Not susceptible to eavesdropping

• Fiber-optic cabling disadvantages:
– More expensive than other types of networking
media
– More difficult and more expensive to install than
any other network media
– Because it is fragile, it must be installed carefully
and protected after installation

Signal DegradationSignal Degradation
• Degradation sources can be internal or external
• When signals degrade over distance, attenuation
results
• Three internal factors can cause attenuation:
– Resistance
– Inductive reactance
– Capacitive reactance

Signal DegradationSignal Degradation
• When the internal opposition forces are combined
and measured, the measure is called impedance
– External forces affecting network signals include:
– Electromagnetic interference (EMI)
– Radio frequency interference (RFI)
– Both types of interference can degrade and corrupt
network signals as they travel through a wire

Ways to Reduce EMI/RFI onWays to Reduce EMI/RFI on
Network CablingNetwork Cabling
• Keep network media away from sources of
EMI
• Ensure that network media is installed
properly
• Use shielded cabling
• Use repeaters
• Ensure that you install high-quality cabling

Horizontal Cabling StandardsHorizontal Cabling Standards
• Horizontal cabling
– The twisted-pair or fiber-optic media connecting
workstations and wiring closets
• Electronics Industries Alliance and
Telecommunications Industry Association (EIA/TIA)
– Defines a set of specifications, EIA/TIA-568, which
covers outlets near the workstation, mechanical
terminations in wiring closets, and all cable running along
the horizontal path between wiring closet and workstation

Figure 4-7: Horizontal cabling

• EIA/TIA-568B
– Specifies that the maximum distance for a UTP
horizontal cable run is 90 meters (295 feet)
– Also, patch cords (a.k.a. patch cables) located at
any cross-section cannot exceed six meters (20 feet)
• In addition to UTP, the following cable types
may be used for horizontal pathways:
– STP – two pairs of 150-ohm cabling
– Fiber-optic – a two-fiber 62.5/125 multimode cable

Wiring ClosetsWiring Closets
• Contain the wiring and wiring equipment for
connecting network devices, such as routers, bridges,
switches, patch panels, and hubs
• EIA/TIA-568 and EIA/TIA-569 standards apply to
the physical layout of media and wiring closets, with
the latter stating there must be a minimum of one
wiring closet per floor
– Furthermore, when a given floor area (catchment area)
exceeds 1,000 square meters, or the horizontal cabling
more than 90 meters, additional wiring closets are needed

Wiring ClosetsWiring Closets
• The main distribution facility (MDF) is the
central junction point for wiring of a star topology
• The additional closets are called intermediate
distribution facilities (IDFs)
• IDFs are required when:
– Catchment area of MDF is not large enough to capture all
nodes
– The LAN is in a multistory facility
– The LAN encompasses multiple buildings

Proximity to the POPProximity to the POP
• Ensure that
main wiring
closet is
close to the
point of
presence
(POP) to
the Internet
Figure 4-8:
Network spanning
multiple buildings

Proximity to the POPProximity to the POP
Figure 4-9:
Network
spanning
multiple
floors

BackboneBackbone
• Backbone cable (sometimes called vertical
cabling) connects wiring closets to each other in
an extended star topology
• EIA/TIA-568 specifies four different options for
backbone cabling:
– 100-ohm UTP
– 150-ohm STP
– 62.5/125-micron optical fiber
– Single-mode optical fiber

Performance Considerations:Performance Considerations:
Connection SpeedsConnection Speeds
• The real capacity of a network is sometimes
referred to as throughput
• Factors affecting throughput include:
– Type of network devices being used on the network
– Number of nodes
– Power issues
– Network architecture
– Other variables

UtilizationUtilization
• Potential causes of high utilization:
– Video or audio streaming/teleconferencing
– Client/server applications
– Host/terminal applications
– Routing protocols
– Routine maintenance tasks
– Broadcast traffic
– Ethernet collisions

UtilizationUtilization
• Solutions for reducing network utilization
include:
– Segmenting a network with connectivity
– Reducing number of services provided on the segment
– Reducing number of protocols in use on the segment
– Disabling bandwidth-intensive applications or
protocols
– Relocating systems consuming the most bandwidth on
the segment

Calculating Bandwidth and ThroughputCalculating Bandwidth and Throughput
• When considering an organization’s
bandwidth requirements, discover types of
bandwidth-intensive communications
conducted on its network
• Transmission time
– Time it takes a file to transfer from one location to
another

Collisions and ContentionCollisions and Contention
• All stations on an Ethernet segment must share
the available connection with each other
– This means the stations contend with one another for
the opportunity to transmit on the wire
• When considering upgrading an existing network,
check the rate of collisions on the network using
a protocol analyzer or other network
performance-monitoring tool

Resource PlacementResource Placement
Figure 4-10:
Resource
placement

Installing TelecommunicationInstalling Telecommunication
ConnectorsConnectors
Figure 4-11:
RJ-45
connector

Figure 4-12:
UTP wires

Figure 4-13:
Jack wiring

• EIA/TIA-568A
– Wiring method used to indicate which colors are
assigned to which pin for UTP cable
• Punch tool
– Used to punch down cable at the patch panel or
RJ-45 wall jack

Patch PanelPatch Panel
Figure 4-14: Patch panel pins

Patch PanelPatch Panel
Figure 4-15:
Patch panel ports
Figure 4-16:
110 punch tool

Cable Testers:Cable Testers:
Wire MapWire Map
• Important measurement a cable tester makes to check wiring sequence
Table 4-3:
Wire map
error
detection

Wire MapWire Map
Figure 4-17:
Crossed pairs
Figure 4-18:
Split pairs

AttenuationAttenuation
• Attenuation is the loss of signal power over
the distance of a cable
• Signal injector
– Puts traffic on a wire so that a cable tester can
measure attenuation and crosstalk
• The lower the attenuation, the better

NoiseNoise
• Alternating current (AC) signal noises are called
oscillations and can alter the digital signals that
computers receive on the wire
• The motherboard and other internal integrated circuits
of a computer use the chassis as their ground
• Faulty AC wiring can also cause problems with
transmissions because the signal reference ground is
the computer chassis and grounding plate
• A transformer steps voltage up or down where the hot
lead originates and the neutral wire is grounded

NEXTNEXT
• Near end crosstalk (NEXT)
– Measure of interference from other wire pairs
• Causes of NEXT include:
– Split pairs
– Too much wire untwisted at the patch panel, jack,
or connectors
– Bends, kinks, or stretches in the cabling

NEXTNEXT
Figure 4-19:
NEXT test on a
cable analyzer

Distance MeasureDistance Measure
• EIA/TIA-568A specifies maximum cable
lengths for network media
• Cables that are too long can cause delays in
transmission and network errors
• Time-domain reflectometer (TDR)
– Cable tester that can detect the overall length of a
cable or the distance to a cable break

BaselineBaseline
• Take baseline measurements to tell how well
the network is performing at a given moment
• Baseline measurements can include:
– Error rates
– Collision rates
– Network utilization

Network ArchitectureNetwork Architecture
• Logical topology
– Describes the way a signal travels in a network,
which is a function of the access method
• Usually a bus or a ring
• IEEE 802
– Covers issues concerning all types of networks
• LAN, MAN, WAN, and wireless

Logical Link Control (IEEE 802.2)Logical Link Control (IEEE 802.2)
• In the IEEE 802.2 specification, the Data Link layer is
divided into:
– The Media Access Control (MAC) sublayer
– The Logical Link Control (LLC) sublayer
• LLC sublayer is closer to software components of the
protocol stack because it controls data link
communications and defines Service Access Points
(SAP)
• MAC sublayer is closer to the underlying hardware
architecture

Logical Link Control (IEEE 802.2)Logical Link Control (IEEE 802.2)
Figure 4-20:
802.2
specification

CSMA/CD (802.3)CSMA/CD (802.3)
• IEEE 802.3 defines the access method used by
most Ethernet networks
• Jam signal
– 32-bit message to all computers on an Ethernet network
that tells all stations not to transmit
• 10BaseT
– Describes an Ethernet network connected by twisted-pair
cable that can support transmissions of 10 Mbps using
baseband (digital) signals

CSMA/CD (802.3)CSMA/CD (802.3)
• 10Base2
– Also known as thin Ethernet
• 10Base5
– Also known as thick Ethernet
• Fast Ethernet
– Also known as 100BaseT
• Gigabit Ethernet
– A more recent addition to the IEEE 802.3
specifications

Token Ring (802.5)Token Ring (802.5)
• In the 802.5 specification, Token Ring networks use
token-passing to keep track of which node is
communicating
• Star-ring
– Network architecture utilizing physical star topology with
logical ring topology
• Nearest active upstream neighbor (NAUN)
• Nearest active downstream neighbor (NADN)

Token Ring (802.5)Token Ring (802.5)
• Active monitor
– Computer in a Token Ring network that is
powered on first and that manages the beaconing
process
• Beaconing
– Fault-detection method implemented in Token
Ring networks

Wireless Technologies (802.11)Wireless Technologies (802.11)
• The 802.11 standard for wireless LANs specifies
parameters at both Physical and Data Link layers of
OSI model
• At the Physical layer, infrared (IR) or spread
spectrum technologies are supported
• At the Data Link layer, 802.11 specifies Carrier
Sense Multiple Access/Collision Avoidance
(CSMA/CA) as the network access method

FDDIFDDI
• Fiber Distributed Data Interface (FDDI)
standard
– Responsibility of the American National
Standards Institute (ANSI)
– Describes a network that can span up to 100
kilometers (62 miles) over single-mode fiber-optic
cabling
– Based on the Token Ring (802.5) specification but
with different limitations

LAN Design ModelsLAN Design Models
• You can choose many different network
design models to implement on your network
• There are two basic designs strategies that are
typically followed:
– Mesh design
– Hierarchical design

Figure 4-21:
Mesh network
design

• Compared to a mesh design, a hierarchical
design:
– Is easier to manage
– Is easier to troubleshoot
– Has improved scalability
– Allows easier analysis

Three-Layer Network ModelThree-Layer Network Model
• Divides a network into three connectivity
layers
• Consists of:
– Core layer
– Distribution layer
– Access layer

Three-Layer Network ModelThree-Layer Network Model
Figure 4-22:
Three-layer
network model

Two-Layer Network ModelTwo-Layer Network Model
One-Layer Network ModelOne-Layer Network Model
• Two-layer network model
– Divides a network into two connectivity layers:
• Core
• Access
• One-layer network model
– Includes WAN connectivity equipment and organizes
a network so that is can be easily adapted to the two-
layer and three-layer design models in the future

Two-Layer Network ModelTwo-Layer Network Model
Figure 4-23:
Two-layer
network
model

One-Layer Network ModelOne-Layer Network Model
Figure 4-24:
One-layer
network
model

Network-Management ToolsNetwork-Management Tools
• The most common network-management tools are:
– Cable testers
– Network monitors
– Network analyzers
Table 4-4: Monitor and analyzer

Table 4-4 (cont.): Monitor and analyzer

• Other sophisticated network-management
tools can be used for daily network-
management and control functions
• These tools typically have three components:
– Agent
– Manager
– Administration system

Simple Network ManagementSimple Network Management
Protocol (SNMP)Protocol (SNMP)
• A Management
Information
Base (MIB) is a
database that
maintains
statistics and
information the
SNMP reports and
uses
Figure 4-25:
SNMP in action

Simple Network ManagementSimple Network Management
Protocol (SNMP)Protocol (SNMP)
• Management tasks include:
– Network traffic monitoring
– Automatic disconnection of problem nodes
– Connection or disconnection of nodes based on
time and/or date
– Port isolation for testing purposes
– Remote management capabilities

CMIPCMIP
• Common Management Information Protocol
• Similar to SNMP in that it uses the MIB to
monitor the network
• Not as widely implemented as SNMP
• More efficient than SNMP because the client
reports the information to the management
device

Chapter SummaryChapter Summary
• There are three basic physical LAN topologies
• These topologies typically involve cable
• The IEEE has defined many standards that
have influenced the way networks are
designed and implemented
• One of the largest contributions from the
IEEE is the 802 standard

Chapter SummaryChapter Summary
• Installing media on a network is multifaceted
project
• Obstructions and EMI/RFI must be overcome
• When implementing a network, you can
choose on of three hierarchical models
• Network administrators use network monitors
and network analyzers to manage a network
on daily basis

Network topology architecture

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Viewers also liked

Viewers also liked (20)

Similar to Network topology architecture

Similar to Network topology architecture (20)

More from Dhani Ahmad

More from Dhani Ahmad (20)

Recently uploaded

Recently uploaded (20)

Network topology architecture

Editor's Notes