Networking
Assignment
1.1 A Thin Ethernet / 802.3 LAN will not operate (at
all, not even a little bit) if the cable is not terminated
correctly. What is meant by “termination” in this
context, and why is it necessary?
According to IEEE 802.3 standards for Local Area Network (LAN),
a Thin Ethernet is also referred to as 10Base-
2 Ethernet. The standard specified that 10Base-2 Ethernet
would use Radio-Grade 58 (RG-58) cable. RG-58 cable is a co-axial
(coax) cable and 802.3 specified that the coax cables for
10Base-2 Ethernet were to be connected in a linear bus configuration.
The nodes are connected with the aid of British Naval Connector
(BNC) T-connector.
Since Thin Ethernet uses coax cable on a bus network, it is
vitally important that all open ends are terminated correctly.
The term “termination” in this context means the
removal of the signals from the bus at the open end of its
linear connection by grounding.
It is for this reason that BNC terminators are attached to
the open ends on Thin Ethernet with one end grounded (terminated).
The sole purpose of terminators is to prevent signals from
being bounced back along the cable and thus inducing unnecessary
interference to the original signal. On 802.3 Thin Ethernet,
the coax cable used is RG-58 50-O, so 50-O terminators should
be used on each end, one of which is grounded (as shown in
Figure 1 below).
Figure 1: Connecting a computer to thin Ethernet
Source: http://www.ethermanage.com/ethernet/10quickref/ch4qr_2.html#HEADING1
1.2 Why does the 802.3 standard limit the length of a single
“segment”?
802.3 specifies the physical media and the working characteristics
of Ethernet. For all bounded media, signals (either electrical
or optical) are sent through cables, so it is inevitable that
the signals willsomewhat
degrade along the cable due to the lost of energy and environmental
interferences. Hence, at a certain length, the strength of
the signals will fall below a critical limit and become overwhelmed
by the noise.
It is therefore important for one to understand these physical
limitations when planning a network structure. The 802.3 standard
sets out the length limit to provide guidance for ensuring
these limitations are not exceeded.The 802.3 standard specifies
the maximum segment length based on the signal attenuation
of a particular media in terms of decibel (dB – a logarithmic
expression of the ratio between two signals power) loss per
unit length (for each meter) of cable. Cables may work beyond
the limit distance the standard specified but usually with
severely degraded performance.
1.3 What are the lengths limits for each 802.3 medium
(Thin Ethernet, Thick Ethernet, Cat 5 and Fibre)?
The length limit 802.3 specifies for different mediums are
shown in the table below:
| Type |
Media |
Max. Segment / m |
Max. Network / m |
| Thin Ethernet |
RG-58 Co-axial cable |
185 |
925 |
| Thick Ethernet |
RG-8 Co-axial cable |
500 |
2,500 |
| Cat 5 |
Cat 5 cable |
100 |
No Max. |
| Fibre |
Mono-mode Fibre |
2,000 |
Depends on application |
| Fibre |
Multi-mode Fibre |
400 |
Depends on application |
Table 1: Length limits for each 802.3 medium
1.4 Why is it regarded as preferable to use twisted
pair Ethernet, with its added extra expense of a central hub,
compared to 10Base2 (thin wire) Ethernet?
There are many reasons owing to why twisted pair Ethernet is
regarded as more preferable over 10Base-2 Ethernet.
Twisted pair Ethernet is arranged in a star topology configuration,
so it is more robust than the bus topology configuration in
10Base-2, in which a fault at a single point will bring down
the entire network. Star configuration is also easier to maintain.
For example, adding a new workstation to a segment of the network
will no longer require the entire segment to be taken off line.
It also makes faults finding in cabling much easier to pinpoint.
Another advantage is the huge difference in the number of nodes
allow per segment. A 10Base-T network can accommodate up to
1024 nodes, compare with 90 maximum nodes on a 10Base-2 network.
More importantly, twisted pair Ethernet such as 10Base-T uses
RJ-45 jacks, which is a lot easier and user-friendly than the
BNC connectors. Most people are familiar with wiring twisted
pair cables because of its similarity to telephone cables with
RJ-11 jacks. It also makes the administrators’ job easier
by taking terminators out of the equation.
The potential of achieving bandwidth of 100Mbps (on 100Base-TX)
on twisted pair (Cat 5) networks has ultimately outperformed
the 10Base-2 Ethernet. Gigabit Ethernet has also been made available
recently on twisted pair.
2. You want to connect
two networks that are both 10BaseT Ethernet networks. The two
networks are located in separate buildings so that they are
approximately 1,000 metres apart.
Which of the following network media would you recommend?
• Co-axial cable
• Single mode fibre
•
Multi mode fibre
• Microwave
• ISDN
Why did you choose the above media, giving full reasons of advantages
for this choice against the alternatives?
For this particular application, the length is obviously the
most important consideration. From the choice of media above,
only Single-mode fibre, Microwave and ISDN can be used to connect
the two networks. A thick co-axial cable (RG-8) segment cannot
exceed 500m in length, so assuming it is not possible to use
a repeater, it will not provide the length to connect the two
together. For the same argument, one may conclude that Multi-mode
fibre is not appropriate since multi-mode fibre link cannot
exceed 400m in length to guarantee proper signal timing. Although
special equipment such as Xpresso 16XX CD may be used to extend
the range of the Multi-mode fibre to 2000m, it is not the ideal
solution comparing to single-mode fibre since multi-mode utilises
a mixture of wavelengths and the equipments use are more complex..
Apart from the consideration for distance, the bandwidth of
the network must also be taken into account. The two networks
are currently running 10Base-T Ethernet, so the maximum rate
is 10Mbps. With ISDN at Primary Rate Interface (PRI), the maximum
bandwidth of the link will be limited to 1.544Mbps. Hence, it
will create a bandwidth bottleneck situation, so it is not recommended.
The only practical options are single-mode fibre and microwave
connection. Microwave is very complicated to set up and it is
very expensive. Although it has the advantage that equipments
can be re-use if the offices are moved to different locations.
If there are obstacles between the two buildings such as a river,
etc. that makes cabling difficult, microwave will be the first
choice for recommendation. However, since there is no mentioning
of such obstacles, it is assumed that cabling is possible between
the two buildings.
If cabling is permitted, then single-mode fibre connection will
be the best suited solution, since it provides a possibility
for the network to be later upgraded to 100Base-TX or 1000Base-ZX
network. Even with a 10Base-F network, which limits the length
to 2000m, there should be sufficient fibre length to go around
the adjacent buildings and connect to the node inside the building.
Nevertheless, the cabling process is the costly in terms of
time and equipment cost, and it may compare to the cost of installing
a microwave network. It should also be remembered that pinpointing
breakages in fibre optic connections is difficult, and replacement
cost may be significant.
In conclusion, the first choice of recommendation is Single-mode
fibre (assuming there are no restriction in cabling), followed
by Microwave connection. Microwave connection is preferred only
if physical connection between buildings is impossible and extra
bandwidth is not expected.
3. Discuss the advantages of using standards when
planning a network infrastructure.
There are a number of advantages for introducing standards
for network infrastructure.
Before standards were introduced, device manufacturers used
to make devices according to their own sets of standards.
Thus, one can imagine if all manufacturers make different
elements of the network without the use of a shared industrial
standard, then it is highly improbable that two devices made
from different manufacturers will work with each other. At
that time, almost all of the equipment used on a network must
be supplied from the same manufacturer to ensure the network
operates. Since standards were introduced, all the manufacturers’
efforts were channelled together to make new range of products
that are more compatible with each other. This enhances the
choice of networking devices tremendously, which also increases
the re-usability of the network components.
Standards also ensure the quality of the components. For instant,
the cabling standards ensure that a specific grade of cable
will provide the minimum performance sets out in the standard.
The performance of the connecting media is vitally important
with networks; they are often referred to as the backbone
of the network. Standards allow the cables to be categorised,
so that one may utilise a cable from the category specified
in the standard and be sure that it will not corrupt the passing
data beyond a level the system can tolerate.
Apart from the type and quality of media use, standards also
specify the maximum length of cables, maximum number of nodes
and maximum / minimum distances between nodes for segments
and the network.
Therefore, a network infrastructure that is planned and built
to work within standards and utilise components that meet
the standard specifications will ensure components are of
good quality, equipments are compatible with each other and
media connections are within the limitation of the physical
constrains. Hence, good quality connections can be established.
Such network will be organised and easy to understand, which
reduces the chance of incorrect arrangements and makes fault
finding straightforward. Future upgrade planning for such
network will be relatively trouble-free since all the constraints
in the network are documented and thus all the physical restrictions
are known, which enables the upgrade process to replace only
the incompatible components.
4. Describe the following tests that are used for
in TSB67 and why they are used.
• NEXT
• FEXT
• Wire map
• Length
• Attenuation
TSB-67 stands for Technical Service Bulletin 67. It provides
detailed Transmission Performance Specifications for field
testing UTP & STP cabling systems. TSB-67 defines four
measuring methods (Line Map, Length, Attenuation and Crosstalk)
to categorise cables into standard Categories, such as Cat
3, Cat 5, etc. These methods basically test the cables for
consistency. They are the foundation for certifying twisted
pair cable standards.
4.1 Line Map
Line mapping (also called end-to-end connectivity) is a method
to test the wiring of the twisted pair cables. It enables
testers to detect any incorrect configurations / wiring of
the cable. Unlike the twisted pairs in a telephone system,
each wire is expected to go straight through in a normal LAN
connection. If any of the wires in the cable is shorted or
crossed, some of the signals will be sent to the wrong devices,
therefore causing disruption to the network or even bring
the entire network system to a standstill. Therefore, it is
vitally important that all the cables are checked and certified
to be correctly mapped before used.
4.2 Length
The length of the cables on a network has a significant impact
on the signals. For UTP & STP, the length of the cables
is restricted by the maximum tolerable signals degradation.
It is important that all cables are certified and stated clearly
the exact length, so that application of the cables will not
exceed the desired length. A common instrument used in cable
length testing is the Time Domain Reflectometer (TDR). It
works very much like sonar – it sends out a signal along
the cable, which is bounced back at the other end. The time
difference is used to calculate the length. This instrument
is particularly useful for administrator to ensure the cables
laid inside walls are within the maximum limit. It is also
a great tool for finding the exact location of breakage.
4.3 Attenuation
Attenuation means the overall signal strength loss in the
cable. It is measured in dB. In general, the more attenuation
there is, the weaker the signal is at the receiver. Attenuation
characteristics of UTP & STP cables are critical for LAN.
This is because data signals used in LAN can be of frequency
up to 100MHz, many times greater than the original frequencies
(5 -10 kHz) twisted pair was designed for. So the deterioration
of the signals falls to critical limit in a few hundred feet.
Attenuation is used to specify the standard of the cable so
that each cable can be proven to be “fit for the job”
and transmits the signals through without any lost of data.
4.4 Crosstalk
The term crosstalk originally came from the phenomena in telephone
network where a wire may take the induce signals from the
neighbouring wire and causes one phone over-hearing others’
conversations. Obviously, this phenomena posts security problem
in the Plain Old Telephone Network, it is also vital that
cables on LAN do not perform in such fashion. Otherwise, the
wires in the cable will behave as if they are shorted, which
will disrupt the network. So TSB-67 specifies some tests to
ensure cable quality is high enough to prevent crosstalk.
4.4.1 NEXT
NEXT stands for Near-End Crosstalk, which means that portion
of a signal is induced into the adjacent wire. These induced
signals may be significant and thus falsely identified as
genuine signals. TSB-67 measures crosstalk from a transmitting
pair to an adjacent pair in the same cable sheath. The measurements
provide a way of assessing the tendency of crosstalk to an
adjacent pair.
4.4.2 FEXT
FEXT stands for Far-End Crosstalk, it is basically the same
as NEXT but the measurements are taken from the opposite end.
So the signals will have travelled through the entire length
of the wire along with the induced signals. FEXT is considered
not a consistent test because the amount of FEXT varies significantly
with the length of the cable. “Equal Level” FEXT
(ELFEXT) is used instead, equal level means that the attenuation
effects are normalised, which makes the measurements independent
of the cable length.
-
BIBLIOGRAPHY:
-
http://www.cisco.com/univercd/cc/td/doc/cisintwk/ita/nums12.htm
-
http://www.ethermanage.com/ethernet/10quickref/
-
http://www.etherwan.co.uk/models/EX16xxCD/EX1616CD_EX1624CD.htm
-
IEEE Std 802.3-2002, IEEE Standard
for Information technology –
Telecommunications and information exchange between systems
–
Local and metropolitan area networks – Specific
requirements – Part 3:
Carrier Sense Multiple Access with Collision Detection
(CSMA/CD)
Access Method and Physical Layer Specifications.
-
NASH, Jason. MCSE ACE IT! NETWORKING
ESSENTIALS, IDG Books Inc., 1998.
-
http://www.protelturkey.com/urunler/tm/acterna/pdf/lan-handbook.pdf
-
http://www.searchnetworking.com
|
