Local Area Network Design

Life Cycle of a Network.

1. Feasibility Study.

This stage involves the sub-phases of problem definition and investigation.

The problem definition attempts to find the problems that exist in the organisation that caused management to initiate the study.

The investigation sub-phase involves gathering input data to develop a precise definition of the present data communication conditions and to uncover problems.

Interviews with users to develop a precise definition of current data processing needs and to identify current problems..

Review of technical reports and documents usually reveal a great deal of system procedures.

Questionaries and surveys help gather information.

Possible solutions are found and from those the best solution is know realised

Outcomes are at the end of this stage:

· Findings of the feasibility study.

· Alternative solutions in addition to the best possible solution.

· Reasons for continuing to the next phase of the process.

If a realistic solution was not found, recommendations for another study and the methodology to follow in order to arrive at a feasible solution.

2. Analysis Phase.
This stage uses the data gathered in the feasibility study to identify the requirements that the network must meet in order to have a successful implementation.

Outcomes:

· Network identification and description

· Benefits of the proposed network

· Current status of the organisation and existing networks

· Network operational description.

· Data security requirements.

· Applications available for this network.

· Response Time.

· Anticipated reliability.

· Data Communications load that the network will support.

· Geographic Distribution of nodes.

· Documentation and Training.

· Network expected Life.

Reference materials used in preparing the report

3. Design Phase.

· All components of the network are defined so their acquisition can be made.

· This is usually the longest phase. Reasons are the size of the job and unexpected complications which usually arise. Management can also affect the time for reasons which are mainly economics.

· Internal specifications or “blue prints” of how the network operates, including modules used for building the network are produced.

· External specifications are the interfaces that the user will be required to use are also produced.

Both of these specifications may include data flow diagrams, logic diagrams, product modules, prototypes, and results from the network modeling

· all the network requirements. These requirements should now be put into priority by dividing them into:

Þ Mandatory requirements which are those which will make the network both operational and effective.

Þ Desirable requirements which are improvements for both the network and users. They may come into play if they are warranted.

Þ Wish list requirements are the requirements that everybody would like to have so they can make their job easier or more pleasurable. But always remember the unexpected problems or costs that usually do come up.

· Network hardware components are detailed and itemised and an outline of procedures for installation and testing of the network are produced.

· The final document produced in this phase is the “blueprint” for the remaining phases of the life cycle.

· Several items that designers will have to address during this phase are:

Response Time which is the time that expires between sending an inquiry from a workstation or terminal and receiving the response back at the workstation. The Response Time is calculated by totalling the:

Þ Message being sent time,

Þ Time required to reach the host,

Þ The Host processing time,

Þ The transmission back from the host to the workstation,

Þ Time required by the workstation to display the response back to the user.

The quicker the response time however increases the costs for the system.

Best response times are found by examining other systems which are similar to the new network to be implemented.

Reasons for increasing response times are long polling times, increase in travel time on circuit, Protocol response delays, hardware delays and Host processing delays.

Network Modeling.

performance and then costs. The data making up the model can be placed in specially developed software packages and this allows us to see how the network may behave.

Once the model is created a simulation tool can provide us with a overall performance capacity and cost analysis for each network alternative. This allows us to examine “what if” situations visually before any extra money is spent.

A good model will allow designers to find weaknesses in the plan and anticipate any problems that may appear during installation or operation of the network.

Output:

Recommendations for a particular network design.

Geographic Scope.

Geographical maps of the scope of the network should be prepared after the model is created and fine tuned. It can be Local, city-wide, national or international.

Message Analysis.

This involves identifying the message type that will be transmitted or received at each terminal or workstation. Message attributes are identified including the number of bytes for each message.

Message length and message volume identification are critical to determine the volume of messages that will be transmitted through the network.

Daily traffic is measured into hourly traffic to provide information about any problems found during peak hour traffic.

Software/Hardware Considerations.

The software will determine the operations:

Asynchronous or Synchronous.

Full or Half Duplex.

Speed of transmissions.

And what types of networks which can access the network.

Today’s designers use the OSI seven layer model to formulate the networks protocols. This will allow easier growth in the future.

Hardware required:

1. Terminals.

2. Microcomputers and network interface cards.

3. File servers.

4. Terminal Controllers.

5. Multiplexers.

6. Concentrators.

7. Line-sharing devices.

8. Protocol converters.

9. Hardware encrypting devices.

10. Switches.

11. PBX switchboards.

12. Communication circuits.

13. FEPs (Front end processors).

14. Port-Sharing devices.

15. Host computers.

16. Channel extenders.

17. Testing equipment.

18. Surge protectors, power conditioners, and uninterruptible power supplies.

Final Hardware Considerations

Configurations of software protocols and network operating system need to be taken into account. Results should show the least expensive system that meets the companies requirements.

Finally and most importantly, selecting a vendor to supply all our needs is most important. The vendor should be considered if they can also supply maintenance, provide future upgrades and assist with future expansions.

4. Implementation Phase.
Steps:

Software Acquisition.

If the network is new, necessary network operating system, application software, managing software and communications protocols must be procured. Tender bidding is often used.

Hardware Acquisition.

Hardware procurement can be done from the same software vendor or a third party vendor may be used. Some software can run on a multitude of computer hardware configurations usually called platforms. Deciding on the software often narrows the selection of a Hardware vendor.

Installation.

Assembly of hardware and installation of software is the next stage. The individual components need to be assembled into what will become the new network. The final product of this phase is an operational network system.

Testing.

Hardware and software are both tested at the same time. Maximum work loads should be processed to provide statistics that indicate best and worst possibilities of network efficiency. This allows us to gather feedback so further fine-tuning can be done before it becomes fully operational.

Test plans (test data and expected results) should be performed, this allows us to submit a test to the network knowing what should the outcome be. All possibilities must be tested and any problems met should be rectified and totally retested.

Documentation.

This should be done throughout the whole development of the new system. Reports that document every aspect of the network, from its conception until final implementation, must be present for audit trail purposes and must always accompany the network. These documents can take the form of reference manuals, maintenance manuals, operational and user manuals, and all the reference materials used in the feasibility study.

Switch-over.

1. This is the transferring of operations from the old to the new. The final product of this step is the active working network. The switch-over plan must include milestones to be reached during the transition period and contingency plans in case the new system does not meet operational guidelines. Maintenance Phase.
The network is kept operational and fine-tuned by network operations personnel. Additionally, updates of software and hardware are performed to keep network operating efficiently and effectively.

Products of this phase are change and upgrade requests, updates to existing documentation to reflect changes in the network, and reports and statistics from the monitoring and control functions of the network.