The School Network

"The need for networked applications and instruction in schools today is a fundamental part of the educational infrastructure, just like good lighting and adequate facilities and resources."

(Cisco, undated)

It is possible for a school to connect to the Internet through a single dial-up or ISDN connection, or through a number of such connections, but a networked connections enables Internet resources to be made use of from any computer in the school that is connected to the network. Recent research by RM, a major vendor of ICT equipment to schools, found that network access to the Internet by UK secondary schools has increased from 29% of schools to 75% of schools, but only 6% of UK primary schools have a local area network. (RM,1999).

A network consists of two or more connected computers. When computers are joined to each other in a network, it is possible for them to share resources such as printers, CD-ROM drives and files. The most usual way of connecting computers is cable, but it is also possible to use optical fibre, infra-red, radio or laser signals. The connection to the Internet can be networked, giving all computers on the network Internet access.

Many school computers would be considered obsolete in any other setting. Sometimes it is possible to incorporate such machines in a school network. This chapter focuses on the connectivity choices available to schools for their local network, and the problem of incorporating legacy systems in the network.

Structured Cabling

Structured cabling is the preferred approach to providing a physical infrastructure for a local area network (LAN) in most commercial situations. It is "a low maintenance, flexible system for moving information to any work area in an organisation reliably and efficiently, and with enough growth margin to support future business needs and applications". ( Network Solutions,1997). Commonly it is combined with a fibre optic backbone where the network extends over a distance.

Cable is a physical medium and it has to be laid where computers might need to be connected. In a commercial situation, it is usual to flood wire premises when a new network is installed. This means installing as many connection points as possible subject to the constraints of the building. One way of doing this is by superimposing a 2 metre grid over a ground plan of the premises and installing a connection at each junction. The advantages of this approach are that it ensures sufficient connection points whatever future expansion in business may occur, and it removes constraints on the placement of desks and other furniture.

Few school buildings would lend themselves physically to this approach; they lack the suspended ceilings and conduits of most modern office buildings. Though it is always a recommended procedure to install more connection points than necessary for current needs, school managers would almost certainly question the necessity of so many connection points in the average school, even in a dedicated computer suite. At the very least, laying cable involves an element of disruption. Not only do cables have to be hidden behind removable panels or in special cable trunking, but power points and RJ45 connections need to be provided for individual workstations.

The most usual method for connecting a network is Ethernet offering a 10Mbps data transfer rate. The physical connection uses unshielded twisted pair (UTP) Category 5 cabling. Category 5 is the preferred choice both because it is the industry standard and it can be upgraded to Fast Ethernet, offering a 100Mbps data transfer rate, should higher bandwidth be necessary. Workstations, PCs, Macs, or NCs, are connected to a centralised hub either directly or through a local hub. Servers, computers which provide services to the network, are also connected to the hub. Using UTP Category 5 cable the maximum link between a hub, or repeater, and a workstation is 100 metres.

Some very specific connectivity rules apply to Ethernet, but as schools are not usually involved in the detailed design of their networks, these rules would not concern most schools unduly. However, as UTP cabling permits no more than four repeater hops between any pair of computers on the network, the topology of the network would possibly be affected if a school occupied a number of different buildings on a large site. In such a situation, it would be necessary to consider connecting local clusters of computers to a FDDI backbone, or providing a wireless connection.

Wireless LANs

In the commercial world, Wireless LANs represent a niche market solution in situations where it may be difficult or impossible to lay a cable based network. They have had a poor reputation as they tend to be slower (bandwidth of less than 2Mbps) and less reliable than cabled networks. In schools, they have considerable appeal as they avoid the need for cabling, though additional power points will probably still be required. With some wireless systems it is less complicated to add additional users to the network. New developments in wireless networking are making bandwidths of 10Mbps a real possibility and data transfer rates of 100Mbps have been achieved under experimental conditions. ( Network Solutions, 1996). Appendix C has further details of the technology involved in wireless networking.

The frequencies used for wireless LANs range from the high end of the ultrahigh frequency (UHF) radio band (900 MHz) through the microwave frequencies (2.4 GHz, 5.8 GHz and 18 GHz) and into infrared invisible light (measured in terahertz). The higher the frequency, the greater a signal’s attenuation, or tendency to weaken over long distances, as it hits a physical object. Below 900MHz signals generally propagate well, even through walls and other barriers. The major disadvantage of infrared is that it operates on a line-of-sight basis so that, while signals can bounce off walls and ceilings, they do not travel round corners without assistance and they cannot pass through walls.

There are several examples of schools using wireless connections for part or the whole of their LAN. Xaverian College in Manchester has seven buildings, five of them listed, on different sides of a busy road. An infrared link provides the network link across the road. In this situation, cabling is not a realistic option. (Johnston, 1999). Another, but more expensive, wireless solution in this situation would have been the use of a laser beam, such as that being developed by the Israeli company, Jolt. (Jolt).

Saltus Grammar School in Bermuda is being used by an American company, RadioLAN, to test a 10 MBPS wireless LAN. Each pupil has been provided with a laptop computer to enable them to access the school network wherever they are on the campus. The school web site shows a photograph of all the students in the school hall using their laptops. This network is fully ethernet compatible and uses protocols complementary to TCP/IP as described in Appendix C.

Though the concept of wireless LANs is new to schools, they are already familiar with wireless technology for registration and other administrative functions. For some years Bromcom has been marketing a dedicated notebook computer for these purposes. At BETT’99, the annual trade fair for educational technology, a new version of this product was being demonstrated with a browser interface to enable it to be used on a school intranet.

ICT Suite or distributed network

Schools do have to give serious consideration to the merits of locating their ICT facilities in one or more dedicated computer rooms or providing cabling for a distributed network throughout the school.

ICT plays many different roles in schools. It is a curriculum subject to be taught, learned and examined and a resource to support the teaching of other subjects. Developing ICT competence is increasingly recognised as a core skill preparing pupils for their post-school lives. ICT is not only used in the classroom, but teachers are making increasing use of computers for administration and record keeping. It is used extensively by the school administrative staff. Schools not only have to consider how best to provide for the needs of pupils, but also whether to connect other users to the school network. Before the introduction of school intranets the solution was relatively straightforward; a networked computer room and a few stand alone computers. This is no longer adequate.

Secondary schools tend to centralise ICT facilities more than primary. A popular model is to have a cluster of computers available within different curriculum areas to provide ease of access to resources when they are needed. Primary schools are less likely to be networked, and tend to have one or more stand-alone workstations in classrooms available for use as needed. The nature of primary school lessons is to enable children to work simultaneously on a range of activities around a common theme; some may be doing artwork or writing while others are using the classroom computer.

The EDSI projects were established in 1995 to explore the potential benefits of broadband internetworking technology in schools. In the research conclusions, it was noted that generally it was problematic to have either only a central resource or only classroom machines, and that decisions about the siting of machines were often made at short notice and without sufficient or appropriate guidance: "The centralised computer-suite approach is, in our view, less effective than having fewer computers in a greater number of areas, for example in each classroom or subject area…In schools with a large number of networked workstations, there was a shift towards this more diffused model of distribution.". (EDSI Group A Executive Summary para: 99).

Legacy Systems

Until recently, few schools have had a co-ordinated policy for the acquisition of ICT equipment. As Owen Lynch, Chief Executive of BECTa said in a BETT99 keynote address: "To date, we have an infrastructure of resource in schools that is varied, often not compatible with itself and unsuitable to the increasing opportunities provided by communications, multimedia and networking. That infrastructure was purchased with varying degrees of knowledge and understanding of the technology and with varying degrees of strategic intent." (Lynch, 1999). These outdated systems represent a substantial investment in hardware and software.

A business approach might be to recognise the difficulties presented by incompatible systems and to decide to salvage that which is salvageable, replacing old systems with modern compatible ones. Within schools, not only is there is a reluctance to dispose of anything that works, but there is a belief that the disruption caused to learning and teaching by radical change is considerable.

Teachers want to know whether it is possible to continue using their old machines on a new network. School managers want to know whether the benefits gained from so doing are worth the cost and complications. It is sometimes possible to incorporate older machines on a network through cross-platform networking. An alternative approach is to continue to use the legacy software with networked emulation software.

Cross-Platform Networking

Xemplar, one of the leading suppliers of ICT equipment to schools, assures schools cross-platform working is not only possible but also easy to achieve: "Imagine being able to rejuvenate systems which have sat in cupboards because the floppy disc programs required to work them have been lost… Xemplar’s MATRIX® networking solution guarantees interconnectivity between different, and older, computers to the latest technologies on the market. Xemplar’s move to supply PCs as well as Acorn and Apple hardware and software, means trouble-free connectivity for schools." (Xemplar Education, 1998).

The approach taken by Xemplar is to utilise network or ‘thin-client’ computing. Networks consist of a number of workstations, or PCs, connected to each other. Usually some applications are stored on the network server and some on the hard disk of the workstation. Data may be stored on the server, or on the workstation, or both. In the ‘thin-client’ model, all applications and data are stored on the server. The network computers (NCs) have no internal hard drive, no CD-ROM drive and no floppy drive. They boot from either software stored in read-only memory or from the remote network connection. Older computers can be configured as NCs.

Network computing can be useful in a school setting, apart from offering the possibility of networking older machines. As all the applications are kept on the server, it is almost impossible to install unauthorised software or to tamper with settings. The network administrator, who is frequently the school ICT co-ordinator "has complete control over all the software that can be accessed from the NC terminals and only has to look after the configuration of one machine – the server – rather than hundreds". ( zdnet, 1998).

Protocols

For computers to be networked together, they need to be able to communicate with each other. Getting computers to talk to each other can be a major difficulty in networking legacy systems or computers built on different platforms. Citrix® has developed an inherently platform independent application called WinFrame® which uses a distributed Windows presentation protocol, ICA®. It is conceptually similar to the UNIX X-Windows protocol and uses industry-standard protocols such as TCP/IP. It supports DOS, Windows 3.x, OS/2, Macintosh (Motorola and PowerPC), Windows 95/98, Windows NT, Windows CE, UNIX, Java, RISC OS and ARM.

"During handshaking, the ICA client communicates such information as screen resolution, colour depth and cache size. The protocol then adjusts to this information. This communication allows for a wide variety of ICA clients, ranging from fixed function monochrome terminals to high-end workstations. On the client, users see and work with the application’s interface, but 100% of the application logic executes on the server." ( Citrix, undated).

Connecting Macintoshes and PCs presents fewer technical difficulties. PCs can be fitted with Network Interface Cards (NICs). All Macs have at least one type of network interface – LocalTalk, and more recent models include Ethernet. Older Macs can have an Ethernet card fitted. Although it is possible to connect PCs to an AppleTalk network using, for example a PCTalk Adapter and Personal MACLAN software, the generally preferred protocol for interconnectivity is TCP/IP. In a school situation where there is an existing Macintosh LocalTalk network, it would be less expensive to add a LocalTalk card to a Windows PC than to replace the LocalTalk cabling.

Emulators

Emulation software is available enabling a PC to behave as though it was a different machine. This software is usually used to enable the use of legacy games software, but there is no reason why it should not be used in an educational setting to run legacy educational software. Emulation programs are generally shareware or freeware and can be downloaded from appropriate web sites. Theoretically, there would appear to be no reason why a server should not offer, for example, BBC micro emulation to permit the continued use of legacy software. Whether or not this would be desirable is debatable.

Emulation software cannot be described as ‘plug and play’. Much of it has been written by enthusiasts for enthusiasts and requires a degree of technical understanding for its installation and usage. As most schools with older systems are primary schools, and primary schools tend not to have much technical expertise available to them, this would seem to be less than useful in most school settings. Some sources of emulation software are given in Appendix H.

When developing their ICT plans, schools need to remember that however desirable it might be to retain legacy hardware and software, and however much has been invested in time and money in these systems, it might be more economical in the long term to replace these systems. If necessary the older computers can be retained as stand alone machines. It is simply over optimistic to assume that existing hardware is capable of being upgraded or added to a network.

Giralang Primary School, Canberra

Giralang is a small primary school with approximately 380 students in the northern suburbs of Canberra. It provides an example of what can be achieved with legacy systems in a network.

In 1995 a new IT co-ordinator was appointed to the school and the Headteacher attended a conference about IT. Both decided to place upgrading the school IT resources high on the school agenda. At that time the school had 22 Macintosh LC II (4/40) computers, 2 Mac classics and a small lab of Apple II E computers. The Macs were networked as a LAN using LocalTalk.

The issues facing the school were:

How best to upgrade the facility on an annual budget of $8000
How to increase student access to computers in school time
How to provide Student access to the Internet
How to provide greater security and filter out unwanted web sites
How to professionally develop and entice reluctant staff members into the IT arena
How to reduce the demand on time to find relevant materials on the Net
How to open up new ways of handling administrative matters such as assessments and student records.

An existing Mac Classic II was set up as a Bulletin Board Systems (BBS) server using First Class Communication Protocol (FCCP). This enabled the computers to be used to exchange mail, graphics and live chats. It was also possible to download from other FCCP sites and to upload student’s work.

In 1996 the school introduced Netscape and changed to TCP/IP. WebWhacker was used to download many curriculum relevant sites on to the intranet server. As a local network is faster than an Internet connection, pupils were able to use Internet resources at minimal cost. (Long and Smith, 1997 , and Cant,1997)

Though the NGfL is providing a means for many schools to acquire new ICT equipment, Giralang does demonstrate what can be achieved with minimal resources

Return to Contents