A computer on its own is referred to as a stand-alone machine, and most domestic computer users use their PC’s in this way. However even small commercial organisations may want to link their computers together. Two computers linked together so that they can exchange data form a network. However when we think of networks, we tend to think of lots of machines linked together.
So what are the advantages of linking machines?
Sharing of resources such as scanners and printers;
Common information can be shared.
Easier to install a program on a central computer rather than having to do it individually for each computer.
E-mails can be sent between users.
New users and equipment can be added easily.
Different types of computers can connect to each other.
Disadvantages are:
If a file server goes down, the users cannot access programs.
If the network is not operating properly, hardware and software resources cannot be accessed.
The network’s efficiency depends a lot on the skill of its manager. A poorly managed network (or one where the manager is obstructive) becomes frustrating to use and may be less efficient than stand-alone machines.
Novices might do something wrong and mess the network up.
Hard to keep secure from hackers.
If a network is not well designed, it might not cope with an increase in traffic without a significant decrease in performance.
“The network’s down” is a phrase often heard in work places and it can seriously disrupt the operation of an organisation. Network managers enjoy the security of tenure of a manger of a premiership football team…
Question 1 Assess two advantages and disadvantages of linking computers in a network. ANSWER
There are two kinds of network:
How are local area networks set up?
In large networks the workstations are connected to a central machine called a server. The terminals may or may not be able to act as stand-alone machines as well. In client server architecture, the devices act as clients sending in a request for service to the server, which is a computer dedicated to managing the networks. The files set up by the users are held on the server rather than the individual terminals. However it is possible to save data on individual machines if they have a hard disk. Associated with the file server may be a print server that stores all the print jobs.
Peer-to-peer architecture is used on small networks. The PC’s can communicate with other PC’s without going through a central server. They are used where people do their own work but occasionally need to communicate with others and share data. However, if one of the stations is turned off, data stored there cannot be accessed.
Comparing the two:
|
Type |
Server
based |
Peer
to peer |
|
Examples |
Novell Client |
Windows for Workgroups |
|
Storage |
Central file server |
Storage on each machine |
|
Software held centrally |
Software held centrally and distributed by the server to requests from the individual PC’s |
Software held on each machine. |
|
Security |
ID’s, Passwords, and access levels are centrally controlled |
No central control of security |
|
Back up |
Done regularly by the central system |
Back up is responsibility of users |
|
Reliance |
All users reliant on the server. If it goes down, then the whole network is unusable |
No reliance on a central computer |
|
Communication |
Through the central computer which has protocols to deal with different kinds of computer |
Each computer needs to have software to deal with communication to other computers. |
|
Advantage |
Speed and power |
Less expensive and easier to administer. |
In some networks the terminals can act as stand alone machines; in others the terminals have to access everything from the server. Although these dumb terminals look like a PC, they cannot do anything, except perhaps the entry dialogue.
How are networks arranged? There are three main ways:
Sometimes the computers are not directly connected to the central computer, but to a distribution box called a hub. You can see hubs in offices as glass fronted metal cabinets bolted to a wall, often rather badly. There are boxes inside with lots of wires and little green lights that flash on and off. Failure of one terminal will not affect the others, although the failure of the server will.
A ring network.
Ring networks are normally found with peer-to-peer networks. Messages pass around the ring on a token until a computer picks up the message.
Question 2 Explain what the difference is between a network with a server, and a peer-to-peer network. Suggest what kind of organisation would use each type. ANSWER
If the computer for which a message is intended is busy, the message can end up going round in circles and can be picked up by another. This can be a good thing, but there are security issues.
3. Bus networks share a single common cable, rather like a universal serial bus. The cable has a terminator so that messages are not reflected at the end. One of the proprietary standards is called Ethernet. Another is BNC.
There has to be a way to ensure that all the computers don’t shout at once, otherwise the data will get totally scrambled up. So software is incorporated so that one computer at a time communicates with another or the server. Each packet of information is given an address to which the data is to be sent and for a short time the cable will only carry this information. The server asks each computer in turn if it needs to communicate. This can take a while in computer terms, but for us the process is very rapid. It is important to avoid data collisions. Each station checks that there are no transmissions before it starts to make a transmission. It works well if everything is not too heavily loaded. If a class logs on at the same time, the whole network grinds along very slowly.
Bus networks can be vulnerable to breaks. Suppose we have five computers, A, B, C, D, and E. If there is a break in the line between C and D, computers D and E won’t work. Similarly, if computer C breaks, D and E could be unable to work as well. The advantage of a bus network is that it is cheap to install, as installing lots of cables can be tedious.
A bus network can be used with a server or in peer-to-peer systems.
Question 3. A school installs a bus network, with outlets for terminals in each room. Where a computer is present, it is linked to the terminal with two wires. If there is no computer, the inlet and outlet are linked with a short loop of cable. During a boring lesson, a pupil removes the loop to see what happens. Within minutes he is being bawled out by the Teacher in charge of the Network. Suggest what has happened to the network. ANSWER
Very large networks can have combinations of the connection methods above, with several hubs. In a university, there may be a server within each building and a central large server in the ICT Services office. Each sub-network is connected via bridges to other parts of the network. A gateway connects the network to the outside world. St Mary’s College (where I teach) has a star network.
Question 4 Give two advantages and two advantages of
a bus network ANSWER
The Role of the Network Administrator
All networks need maintenance, otherwise they will grind to a halt if ignored. Even a stand alone machine needs a certain amount of periodic maintenance to keep its performance tip-top. For a PC, this might include:
Back ups
De-fragmentation of the hard disk
Clearing out temporary files.
Network response time is adequate.
The network responds to the needs of its users.
Upgrades to hardware are carried out when necessary.
Software is loaded and upgraded when necessary.
The system is backed up so that data is not lost if the server goes wrong. In a college the loss of data would be very tedious. In a company it would be disastrous.
Security is maintained, to prevent unauthorised access and the placing of malicious programs. On many networks the download of software is strictly forbidden and can be a disciplinary offence.
Networks are sophisticated entities. No longer can the network be entrusted to Darren who knows a bit about computers. There are courses in network management and the network administrator in many organisations has a highly skilled and responsible position. (Sometimes it is not the most popular position as one cannot please all of the people all of the time.)
Question 5. Explain why the network administrator is considered a key employee in many organisations ANSWER
Wide Area Networks
A wide area network (WAN) is used where devices are more than a few hundred metres apart. The computers require extra equipment such as a modem connected to telecommunication lines. These can be:
Public, in which payment is worked out on the amount of time used. The computer dials up a particular telephone number that is charged at local rates. Domestic access to the Internet costs about 1 p per minute for dial-up connections. Many ISPs (Internet Service Providers) have a pre-paid package that enable free access to the WAN.
Private, in which case an annual rental is paid, and the access costs nothing extra. The private lines are dedicated to that use.
The fixed lines can be:
Ordinary wire (twisted pair) used by most telephones.
Coaxial cable, which has a central conductor surrounded by a conducting braid. Interference is a problem with twisted wires; with coaxial cables it is not.
Fibre optic cables in which the messages are sent as pulses of light rather than electricity. The transmission quality is much higher than wires. The speed of transmission is about 10 times faster than copper wires.
Microwaves. These are short-wave radio waves. They travel in straight lines so are limited by the curvature of the earth, so cannot be transmitted more than about 50 km. Mobile telephones use microwaves. It is possible to access any WAN from a laptop computer via a mobile phone.
Communication satellites form an important part of international WAN’s. These depend on geostationary satellites that are 30, 000 km in space, which have an orbital time of 24 hours, so remaining at a fixed point relative to the ground.
Computers are connected to telephone lines by modems (modulator/ demodulator) which are devices that convert digital signals to analogue signals.
The amount of data that can be sent depends on the bandwidth of the line. A normal telephone has a limited bandwidth, which seriously restricts the volume of data that can be transmitted. So an ordinary Internet connection is limited to 56 000 bits per second and is often slower than that. It doesn’t matter that much for an ordinary e-mail message, but when you are downloading big files, it can take for ever and a day.
The Integrated Services Digital Network (ISDN) has a much wider bandwidth and allows a greater volume of traffic, voice, computer, and video data. All of these are encoded in digital form. Therefore it is much faster than an ordinary telephone wire.
Speed of transmission can also be increased by compressing the data in the form of zip files. Repeated bytes are replaced by one copy together with how many repetitions there are. Quite large compressions are possible, up to about 50 %. You can get more compression with a word-processor file (up to 80 %). Also if you zip up several files together, the overall file size is reduced.
We can move data in synchronous or asynchronous transmission:
PC’s use asynchronous transmission in which each character is preceded and succeeded by a start and stop code. For good measure a parity bit is thrown in. It is fast and economical for small amounts of data.
Synchronous transmission uses the clock signals in the computer, so stop and start codes are not needed. It is less error prone.
When networks communicate with each other, a strict set of rules called a protocol is needed. The rules cover:
Physical connections;
Cabling;
Modes of transmission;
Data format;
Speed;
Error detection and correction.
All equipment with the same communication protocol can be connected together. If the protocols are different, then the data could get disrupted. This can be overcome by translation devices called protocol converters, which enable PC’s to be connected to a mainframe. It prevents compatibility problems such as:
Synchronous and asynchronous transmission.
Character representation with different codes, such as ASCII and EBCDIC. The latter is used in many mainframes.
Different error detection and correction methods.
Question 6 Two people are working at home on a new piece of software. They frequently need to send data and files to each other by e-mail. They live in an area that does not have Broadband. Explain two ways in which they can reduce their transmission charges. ANSWER
Presentations:
Now try the Topic Quiz