Title: NETWORK DESIGN (REKABENTUK RANGKAIAN) TJ3054
1NETWORK DESIGN (REKABENTUK RANGKAIAN) TJ3054
- LECTURER KHAIRUL NAJMY ABDUL RANI
- ROOM NO STM 201
- OFFICE PHONE NO 04-928 3724
- E-MAIL ADDRESSES najmy_at_e-web.uum.edu.my
- OR
-
khairul_najmy_at_yahoo.com - HOMEPAGE http//www.angelfire.com/mo3/najmy
2Overview of computer networking
- What is a computer networking?
- An interconnected collection of autonomous
computer. - 2 or more computers are said to be interconnected
if they are able to exchange (transmit/receive)
information. - Autonomous means that acting independently.
- What is a distributed system?
- The existence of multiple autonomous computers,
which are not visible to the user. - The distinction between CN and a DS lies with the
SW (NOS) rather than with the HW - Why does a computer networking necessary?
- For organizations Resource/Info. sharing, Saving
cost, Communication medium (audio/video
conferencing/net meeting), high reliability,
security, scalability etc. - For society Remote access, Person-to-Person
communication (e-mail/messenger/chat),
audio/video conferencing etc.
3Overview of computer networking
- Computer networking can be classified based on
type of transmission technology and scale. - 2 types of transmission technology, which are
- 1 Broadcast Network 2 Point-to-Point
Network. - Broadcast Network have a single communication
channel that is shared by all the network
devices. - Broadcasting A packet is transmitted, received
and processed by each network machines/devices. - Multicasting A packet is transmitted, received
and processed by a subset of the network
machines/devices (e.g. stmnet, studnet etc.) - Point-to-Point Network consist of many
connections between individual pairs of machines.
4Overview of computer networking
- Computer networking scale
- Local Area Network (LAN) Privately owned
networks within a single building or organization
or campus of up to a few kilometers in size. (10
m to 1 km inter-processor distance). - Metropolitan Area Network (MAN) A bigger version
of LAN and normally used similar technology
covering a group of nearby corporate offices/ a
city either private or public (up to 10 km
inter-processor distance). - Wide Area Network (WAN) Spans a large
geographical area, often a country or continent
(100 km to 1000 km inter-processor distance). - Wireless Networks Apply antennas, telephone with
analog modem and Cellular Digital Packet Data
(CDPD) service. Emerging technology Bluetooth,
GSM and GPRS. - Internet a collection of homogonous
interconnected networks.
5Overview of computer networking
- Vital CN hardware devices includes host
computer, front-end processor (FEP),
modulator/demodulator (modem), and terminal. - Host computer Center computer/processor within
a large data communication system. It can be a
mainframe or a minicomputer or a microcomputer. - FEP is a special purposed computer, which can be
programmed to control and process data in a
network. It can be either a non-programmable and
hardwired communication control unit OR a
programmable device operating as a communication
input/output and processor. FEP components are
channel interface, SW, and line interface unit.
FEP purposes are offloading certain
processing/controlling data from host computer to
a programmed processor, comm. line control, error
control, multiplexing, converting protocol/code
and etc.
6Overview of computer networking
- Modem transforms digital electrical pulse
(generated by computer/terminal) to analog signal
before being transmitted via analog voice grade
circuit and vice versa. - Terminal is a input/output HW located at the
communication circuit end points. It can be
microcomputer or workstation or video terminal or
teleprinter terminal or dumb/intelligent terminal
and etc. - Other necessary HW devices that make network
becomes faster, more reliable and more secure are
MUX, protocol converter, HW encryption devices,
intelligent controller, line adapter and etc. - HW equipments that are commonly used in LAN are
server, repeater, bridge, router, gateway, and
cabling.
7Overview of computer networking
- Several issues have to be touched when discussing
network SW, which is highly structured. - Protocol hierarchies
- Network is arranged on the layer basis. Every
layer will give services to its upper layer.
Every layer communicate with its peer layer (same
level) at other computer. In this case, a
protocol (an agreement on how peer communication
should be made) is required. - Interface and Services
- Entity Active element in each layer.
- SAP The point where service can be obtained.
- Types
- Connection-oriented Establish, use and release.
- Connectionless Each packet is routed
independently. - QoS
- Reliable (acknowledged) OR Unreliable (not
acknowledged)
8Overview of computer networking
- 2 reference models Open Systems Interconnection
(OSI) and Transmission Control Protocol/Internet
Protocol (TCP/IP). - The OSI deals with connecting open systems (open
for comm. with other systems). - The OSI has 7 layers
- Physical mechanical, electrical, functional,
procedural - Data link error flow control intermediate
node - Network routing, switching, network addressing
- Transport error flow control, packetizing
end-to-end - Session establish communication
- Presentation format, compression, encryption
- Application application protocols
9Overview of computer networking
Aplikasi
Presentation
Session
Transport
Network
Data Link
Physical
switching
Electrical voltage
Format data
Hantar mel
physical port
Mod dialog
Flow control
Signal mana hantar dulu
addressing
encription
compression
Berapa laju hantar electrical voltage
Error control
10(No Transcript)
11Overview of computer networking
- The OSI principles
- Each layer can be implemented independently
- Each layer communicates with its peer layer (same
level) - Each layer needs service provided by bottom layer
esp. related to primitive functions. - In the real world
- There is almost no OSI product.
- In LAN, the 2 most bottom layer used LAN
standard. - In WAN, the 3 most bottom layer used X.25, ATM
and FR
12Overview of computer networking
- TCP/IP has 4 layers
- Network interface/physical communication
Connects the network using some protocol so it
can send/receive IP packet over it. - Internet (IP) Permits hosts to injects packet
into any network and have them travel
independently to the destination. This layer
defines an official packet format and protocol
known as Internet Protocol (IP). - Transport (TCP/UDP) Allows peer entities on the
source and destination hosts to do conversation.
Transmission Control Protocol (TCP) is a reliable
connection-oriented protocol that allows a byte
stream originating on one machine to be delivered
without error to any other machine in the
Internet. User Datagram Protocol (UDP) is an
unreliable connectionless protocol that do not
want TCP sequencing or flow control.
13Overview of computer networking
- Application Contains all the higher level
protocols, such as virtual TELNET, FTP, SMTP,
POP3, DNS, HTTP and etc.
Application
NetBios
NetBuei
Physical network
Application
Presentation
Session
Transport
Network
Data Link
Physical
Application
TCP/UDP
IP
Physical Communication
SNA IPX/ SPX Apple-Talk
OSI
TCP/IP
MS Net
14Overview of computer networking
- Standard can affluence the current network design
and implementation, such as tackling problems
related to HW and SW. - Standard are usually enacted/regulated by
vendors, users and ISPs OR NSPs. - Standard are usually drafted by certain
committee, which some of the members are vendor
representatives - Standard can be either de jure or de facto.
- de jure is being enacted by formal/official
bodies, such as IEEE, ISO, ANSI etc. - de facto is being established by informal
organization, such HW/SW manufacturers like
Microsoft, IBM etc. - ISO is being represented by 90 countries dealing
with a lot of standard subjects.
15Intro. of system approach in network design
- Actually, there are a lot of approach types can
be made in designing a new network or in
upgrading an existing network. - In this course, the system approach is being
chosen to design a network or to upgrade an
existing network. - The system approach can be used by network
designer in 1 planning a new data
communication network 2 troubleshooting or
upgrading the performance of a current network
3 planning a public data network. - In designing a network, all the affluence factors
and limitations/problems should be clarified and
highlighted their effects on the network
performance. - The key of success starting from a good and
effective planning particularly, based on
fulfilling the system user interface requirement
16Intro. of system approach in network design
- Currently, network designer usually emphasizes on
user application system. In this case, 2 types
of network user should be classified, which are
organization management (e.g. manager/director/emp
loyer) and individual user (e.g.
employee/subordinate). - Manager needs a reliable network so that they
believe whatever messages that they receive. - User needs a consistent/stable network so that
they can keep maintaining/upgrading their
productivities. - There are 13 basic steps used in network design.
- Feasibility Study
- The main purpose is to improve the current
network or to build entirely a new network for
organization/public.
17Intro. of system approach in network design
- Feasibility Study (Continued)
- The responsibility is to define clearly the
problem existed and make it as (a) written
document. In the document, highlight the
problems, possible root causes, possible counter
measures (solutions) that can be offered and the
objective/benefits of building a new network.
Besides, state also the complete and realistic
costs for a newly designed network or accurate
costs/benefits comparison for an upgraded
network. - Prepare a Network Design Plan(s)
- The main purpose is to provide a framework(s) for
the network design plan. Ensure that the plan
meets the requirements. Identify also possible
constraints. Besides, enlist functions that
newly upgraded/designed network can perform.
18Intro. of system approach in network design
- Understand the current/existing network
- Try to know fully attributes of existing network
system, such as operational functionality and HW
(CPU type, memory size and configurations), SW
(type applications and network drivers/NOS),
protocols and standards, and operational support
methods. Besides, enlist some proposed network
application for future use. If possible put
written ideas in terms of notes/DFD/flowchart. - Define the network requirement
- The purpose is to determine the what in the
development of a new network and to describe
attribute the network must contain.
Conceptually, identify inputs, such as user
needs, organizational needs, existing systems,
regulations, and environment politics in order to
draw out the network requirements and
constraints. In details, review the org. long
and short term plan.
19Intro. of system approach in network design
- Also, simulate the circuit capacity, reliability
and processing time (response time/queuing
theories). If possible, highlight future
requirement(s) that may be uncovered for now. - Define the geography scope
- The purpose is to identify the physical location
that must be interconnected by the newly designed
or upgraded network. There are 4 basic levels in
mapping the network, which are International,
Country, State/City, and Local Area/Facility. - Analyze the message
- The purpose is to identify msg. char., no. of
msg. and how to estimate volume of message. This
step can be done simultaneously with step no. 5
above. The output is a list of msg. With no. of
char. per msg., network link traffic table.
20Intro. of system approach in network design
- Calculate the network traffic and circuit loading
- The purpose is to calculate the circuit capacity
that is needed for traffic based on no. of char.
per msg. no. of msg. delivered per given period
of time. Therefore, needs to calculate the
response time / turnaround time. RIT MIT APT
MOT. Besides, it is possible to show the
circuit capacity req. for each link. The output
is the documented of traffic analysis (msg.
total of char. per day). - Identify network control and security
- The purpose is to control and protect info. from
various threats, such as errors omissions, msg.
Loss/damage, disasters, poor error handling,
viruses/worms, unreliability etc. Use ctrl
spreadsheet to identify, prepare document and
re-evaluate all the ctrl point within a network.
The output are ctrl spreadsheet, list of threats
def. list of ctrl methods performed.
21Intro. of system approach in network design
- Design network configuration
- The purpose is to configure the network based on
network goals/objectives.The ckt. config. is done
btw. The user and host computer. The result
depends on node/terminal and SW/HW changing. The
output is the network config. maps/layouts and
list of network goals. - Consider the SW selection
- The purpose is to overcome the limitation
resulted by terminal usage and diff. types of SW
used thru the using of protocol
converter/router/gateway. The output is the
details of network config. Either it is PTP or
Multiplexed or Multidropped or Packet switched or
combinations of above. Besides, there should be
a description of protocol selected as well as
WAN/LAN SW documentation.
22Intro. of system approach in network design
- Consider the HW selection
- The purpose is to get all the HW required. The HW
of network can be categorized as transmission
media and internetworking devices.The output is
the layout and the config. (multidrop/multiplex,
etc.) showing all the linked nodes. - Calculate the cost required
- The purpose is to estimate the cost of
building/upgrading a network and to review the
complexity of costing on the chosen network
configuration. A lot of techniques can be used
to estimate the cost, such as network cost
analyzer, cost/benefit analysis, voice grade
leased circuit cost, dial-up circuit cost,
wideband analog circuit cost, digital circuit
cost, T-carrier circuit cost, pkt sw. cost, HW
cost, and further design cost.
23Intro. of system approach in network design
- Calculate the cost required
- The network cost generally can be classified into
two groups, which are direct cost and indirect
cost. Direct cost includes processor/server,
comm. devices, common carrier line tariffs,
SW/NOS, spare part cost, and maintenance cost.
Indirect cost will be personnel training and
disruption of normal activities. The output is a
report of network expenditure generated by
network cost analyzer. - Implement the new network
- There will two subtasks, which are selling the
proposed network to mgmt. or to client/user and
implementing the network. Selling the network
can be done orally to top mgmt. There are 4Ps
techniques in implementing the network that are
Pilot, Phased, Parallel, and Plunge.