ITWALES.COM - Lecture 1 - Types and uses of Computer Networks

Lecture 1 - Types and uses of Computer Networks

Table of contents

1.1 Computer Networks Basic Terms

2 Main Uses of Computer Networks

2.1 Business Use

2.2 Scientific Use

3 A brief history of the Internet

1 Computer Networks

Computer Networks have rapidly become an integral part of business life. They allow a business to streamline the overall operation and to interact with suppliers and customers very efficiently.

In many cases computer networks are seen as the solution to every problem within business, and - to some extent - the magic wand to tackle all problems with.

Whilst it is not necessary to understand fully the technical details, it is necessary to understand in full what precise purpose the introduction of a computer network into an organisation serves. This is crucial for the project to be accepted by staff and to help the technicians to design the network accordingly.

This lecture shows the various uses of computer networks, starting to describe the most general types and specifying possible uses in more specific circumstances.

1.1 Computer Networks Basic Terms

Computer Networks

A computer network is defined to be an interconnected collection of autonomous computers. Interconnected means capable of exchanging information, and autonomous means, that there is no dictatorship. The connections allow users to exchange data, and pool and share resources, including

Printer

Data

Computing Power

In contrast to Computer Networks, where the existence of several separate machines is usually hidden from the user, Distributed Systems are computer networks with special software that makes the existence of several computers visible to the user. Distributed systems are usually used in conjunction with parallel computing (see the section "Scientific Use").

Computer Networks are classified according to the distance between individual computers that are attached to the network. The classification includes the following:

LANs - Local Area Network, a computer network that spans a distance of tens of metres at the most. LANs are very common in offices but can also connect several offices together. There is no limit as to how many computers are networked for the network to be classified as LAN.

WANs - Wide Area Network, a computer network that spans hundreds of metres to a few kilometres. Typically the campus spanning network that connects different departments in any University or larger company is called a WAN.

MANs - Metropolitan Area Network, in essence a computer network that is meant to span a whole metropolitan area. This concept was very popular in the early 1990s when various cities envisaged establishing such networks. In practice these networks are very rare, and today only the concept survives. This is mainly due to the increasingly private providers that develop networks as opposed to cities and councils taking the development of such networks into their own hands.

The above classification is purely based on the size of the area that the network covers. In addition to the area criterion other names are common.

Internet - an Internet is a collection of interconnected networks. Quite simply, if two separate computer networks are connected we speak of an Internet.

The Internet - the term "The Internet" is used to describe the network of computer networks installed across the globe. They are interconnected and most computers connect to it via the Internet Service Providers.

Internet Service Provider (ISP)

This term defines a company (usually) that acts as the interface for users to connect to the Internet. The ISP has high bandwidth connections with other computers that are connected to "the backbone" of the Internet.

Bandwidth

Computer networks transfer data. The term bandwidth describes the amount of data that is transferred over any unit of time. It is usually measured in Bits Per Seconds (Bps). To convert Bps into the more familiar measure of Kilobytes per Second (KBps) requires the division of the Bps figure by approximately 8000 (there are 8 bits in any one byte and the prefix "Kilo" means units of 1000). The term Bandwidth is usually used in conjunction with data transmission via any kind of connecting media.

The average speed of data transfer - the bandwidth - depends always on the slowest link in a chain between computers. Traditionally the data transfer rate decreases with the distance between the computers that exchange information. With the advances of technology this difference becomes blurred and less significant.

2 Main Uses of Computer Networks

The main purpose of computers is to enable people to exchange data and information. This can take the form of electronic mail (e-mail) or Intra and Extranets. It is the use of these resources that can decide whether a computer network is an asset or a problem for the operators.

On the most basic level computer networks are great to share resources, such as printers and storage space. On the most advanced level computer networks can carry video data for remote video conferencing for example.

2.1 Business Use

In the business context computer networks can be a real asset. This includes storage and retrieval of information, reducing the need for paper traffic and record keeping in paper format, as well as rationalising the time that staff spends producing correspondence and accounts.

The big danger is that the real purpose of a computer network is not fully assessed resulting in an installation and maintenance plan that is not adequate for its use.

In addition, if the installation of a new computer network is not fully supported by the staff, then the network itself is may not be taken full advantage of, and record keeping becomes a mixture between digital and paper based records.

In a business context the following uses are typical. They are listed in ascending order of complexity:

Share resources such as printer.

Share data on a central server, where all data is collected.

Communicate via e-mail.

Exchange information via an internal network.

Allow all staff access to the Internet.

Use the network and the Internet to allow staff access to business data from anywhere in the world

Integrate the whole business operation into a networked operation, including sales activity, stock holding, quotations, ordering raw materials, control the production process, process invoices, process all the accounts, analyse business performance, quality control.

Each stage introduces an extra layer of complexity to the operation and an extra level of staff expertise and acceptance. In addition to the sheer network installation each level of functionality is combined with a suite of software that needs to be configured once and kept running subsequently.

One example of how complicated things become on the simplest level is access and sharing of data. A detailed plan and assessment of who should have access to what kind of data, and who can change and delete data needs to be drawn up and implemented. From a technical point of view this is very simple, from an organisational point of view this may be a serious difficulty.

With the adoption of a new procedure within the business and the learning of the new processes by the staff, the network becomes ever more mission critical. This means, that if a problem occurs, the operation of the business could be seriously disrupted. To avoid major losses, for example, consider the introduction and purchase of adequate data backup procedures.

And this also means that careful planning is required to make sure that the necessary maintenance of the system is always kept up. Usually the cost of the system and ongoing maintenance is offset by the productivity gains of the improved operation of the business.

It is strongly advisable to make a full assessment of the cost and benefits that the introduction of the new procedure entails and to argue the introduction of a new process in terms of a strict business case.

Very often such undertakings fail because of a communication gap between the technical personnel and engineers who carry out the installation, and the management and staff of the company that takes such a new system on.

Overall, especially in Small to Medium Sized Businesses, the strict assessment of cost and benefit is not carried leading to serious problems.

There is also the tendency to overcomplicate the procedure in software terms. Software can in theory do anything you like (as long as it is an algorithmic process), but different off-the-shelf packages may not interact with each other very well. It is impossible to foresee problems of this kind without the detailed input by experienced engineers.

NOTE : Software and Computer Networks can NEVER be a SOLUTION, they can be instrumental in assisting in the improvement of the overall business operation. Software and Networks need people to work and produce results. And everybody involved will need to know and understand the aim behind the introduction of the new process.

2.2 Scientific Use

Within Science computer networks were used very early on to share resources and exchange data. This goes back to the beginnings of the Internet, and in fact, the Internet as we know it today was created within a scientific laboratory.

It is standard procedure to use networks to share data and use remote computers to carry out large computations. This includes the use of supercomputers in various locations. As an example consider the use of a supercomputer located in Norway by researchers in the UK.

To the operator the physical distance is immaterial. Data nowadays can be transferred very quickly from remote computers.

In addition to the "outsourcing" of computing power, the old system of using one powerful machine within a department to carry out large computations, and many small hosts for the creation of reports is very common.

Research is underway to examine the way in which networked computers can be used for the parallel processing of large computations. The ultimate aim is to produce a system that would use all the resources available on a computer network (and even the Internet) as a computing resource.

The use of computers connected via a network for parallel computations is a non-trivial problem. Many researchers are developing procedures for data transfer and data handling, as well as the adaptation of algorithms for given tasks to parallel processing.

Parallel computations are also carried out by large machines that contain more than one processor. There are two types:

Processor with local memory

Processors with shared memory

Conceptually the difference between parallel computers and computers on a network is very small. Both exchange data, both have separate memory per processor.

Here we now need to make a distinction between processor and computer:

A computer is the combination of a processor, memory and peripheral devices.

A processor is an integrated circuit chip in which the processing is takes place. Without peripheral devices a processor cannot work.

For scientific use it is inconceivable nowadays not to be using a computer network. Interestingly, each laboratory has a technician who is dedicated to the maintenance of the computer network and the safeguarding of data through regular backups. Businesses ought to take notice of this.

3 A brief history of the Internet

The Internet as we know it today grew in essence out of ARPANET. ARPANET was a network of computers in 1969 designed to allow researchers to exchange textual information about advanced research projects within the US Department of Defence. The technology and particularly the protocol stack TCP/IP was developed alongside the needs of the evolving network.

In the late 1970s and 1980s the use of the Internet in form of e-mail and textual information exchange extended to include academic institutions. The creation of JANET (Joint Academic Network) in Britain illustrates this move, when all Universities obtained free access to an important backbone of the Internet.

More and more services and search directories were created, such as Veronica and Gopher, for document exchange in textual and digital format.

In 1990, in the research laboratory CERN in Geneva, the Internet in its present form with hyperlinks and graphical "pages" was developed within a research group that needed to communicate with remote members of the research group in a way which allowed pictures and fast referral to other documents. This involved encoding the information in a text based format with "tags" that could be interpreted by a standard client application independently of the operating system of the client - the browser.

Since then the growth in terms of users has been exponential. In the late 1990s the power of the Internet begins to sink in, and millions of households go online. The impact on day to day life becomes more obvious, as people start shopping, communicate and look for services through the Internet. With the availability of the Internet on television, this tendency will become more extreme still.

The development cycle of technology and new features on the Internet is extremely rapid. An "Internet Year" is said to pass within merely 3 months. In the late 1990s more and more Internet startup companies are valued at astronomical prices on the stockmarket, creating a strong incentive for newcomers to try out and develop new ideas and technologies.

Most of the Internet uses the protocol stack TCP/IP (see below). This technology is being developed by the Internet community at large and details of the latest developments can be found on the Internet site of the World Wide Web Consortium founded by MIT and CERN (http://www.w3.org). New developments include the updating of the "Markup" language for the encoding of the information as well as the development of new "standard" features for the Browser application needed to view sites and interact with the Internet.

As the amount of data and information available on the Internet grows exponentially, so too is it important to be able to search it all in a meaningful way. This job is attempted by "Search Engines", Internet sites that carry information about other sites in a searchable way. Such search engines include Altavista, AskJeeves, Google, Yahoo (and many others) and have become a focal point on the Internet. In addition to Search Engines there are also "portal" sites, allowing access through few "clicks" to areas of specific interest. Offline marketing, i.e. informing potential users in conventional ways of the existence of sites of interest, is becoming a crucial ingredient to ensure the success of an Internet site. This shift from hightech to considering the Internet as a tool for everybody will intensify.

Issues to be dealt with on the Internet include the following. All are made more difficult by the intrinsic international nature of the technology.

Policing of Abuse

Legal and financial implications of cross border trade through the Internet

Protection of youngsters from abusive sites

Intrusion of malignant strangers onto a server

Privacy of data exchange

Non-repudiation, Authentication and Authenticity of messages and data available on the Internet (see chapter 2)

Definition of standards

It is hard to see the Internet not taking the society by storm. And it is hard to overstate the effect it will have on society, with changing habits for social interaction. The breathtaking pace of development of the Internet and its growing availability to the general public in Britain at the end of the 1990s are only the beginning of the development of the Internet. There will still be many more Internet companies and startups, and some will bring real change.

Subscribe to
ITWales Updates
Click Here!