This talk discusses several concepts that have been devised to organize educational networks with common services for universities, schools, and other educational institutions, and some special aspects of operating such networks.
In its early times the Internet consisted to a large extent of universities and other academic institutions. During the recent years it has been rapidly moving towards an open network, largely populated by commercial enterprises as well as private individuals and various organizations.
Universities are facing these changes and at the same time are confronted with the advent of new network applications, a demand for higher bandwidth to accommodate the requirements of imaging and video-streaming, and concepts like digital libraries and virtual universities, for which a fast and reliable network is indispensable. These developments have given rise to plans to enhance the existing networks or to complement them with new high-capacity interconnects.
A majority of (secondary) schools in most countries, however, have been reluctant to connect to the Internet, although its potential usefulness is acknowledged: even if the necessary equipment (computers, routers) can be financed through donations or an increased budget, the access cost for a sufficient bandwidth can be prohibitively high. This is often due to the billing structure of the telecom companies, while bridging the "last mile" with a dedicated cable would also be highly expensive.
Several concepts have been devised or, mostly on an experimental basis, realized to identify possible areas of cooperation between schools, universities, city councils, and such innovative organizational structures like "charitable limited companies" which bundle the interests of individual institutions and conduct the negotiations with service providers and suppliers of equipment.
Some particular problems arise from the operation of a typically heterogeneous technical infrastructure, the variety of schools and their potential, and from the administration of a large number of participants. This situation also demands specific security considerations.
In spite of all the differences between schools and universities the educational institutions have a common interest in providing an open access for students and teachers to resources like multi-media learning materials and interactive applications, and in facilitating an open exchange of ideas.
Extending educational networks beyond the original primary participants of the Internet - the universities - means the inclusion of mostly secondary, but in some cases even primary, schools as well as, for example, adult colleges, vocational schools, training centres, and other institutions for secondary (and tertiary) education (for all of which the term "school" as a generic name will be used in this paper).
Providing Internet access for all these institutions poses a few problems that differ substantially from those the universities have been confronted with so far.
Unlike universities, where usually a great deal of computer expertise is available, it depends on the sort and the size of the school as well as on the individual situation if there are any computer experts among the staff members. Consequently, the amount of advice and counselling that a school needs in order to maintain its Internet connection and to run the basic services, varies considerably; an educational network will have to provide adequate helpdesk facilities for elementary-level inquiries as well as for dealing with the problems of advanced users.
Depending on the type of the school, there will be much diversity in the use of different Internet services. While many schools will only want to present their own WWW homepages, the access to much of the information available through international mailing lists and newsgroups, for example, requires a good command of English and may therefore be restricted to senior students of secondary schools.
In many countries schools notoriously suffer from narrow budgets and therefore cannot afford up-to-date computer equipment without generous help from parents or sponsors. This is the main reason for the heterogeneity of the technical infrastructure which typically can be found at schools. The types of the computers and their age, the amount of available peripheral devices, the operating systems, and the technology of the local area network (if available at all) are often a result of individual donations and "special offers", and they may also depend on the preferences of a teacher who is responsible for the operation and maintenance of the school´s computer equipment.
Since the amount of locally available know-how has grown with the specific software and hardware configuration, it is generally more valuable to keep it up than to standardize the equipment. On the other hand, unusual non-standard installations carry the risk of becoming useless in the absence of a "local guru".
Although most schools have less students than the average university, the number of students at larger schools requires automated procedures for the maintenance of network user accounts. Each user should have his or her individual mailbox, and even if browsing the world-wide web is generally an anonymous act, it is often necessary to identify the local user so that in case of complaints the activities can be traced back to the individual who performed them. Guestbooks, charge-free e-mail accounts, and web-accessible chatrooms are just a few examples of how a web user can send messages to other netizens, and whenever these messages contain obscene or unlawful material the school administration or the provider will be under pressure to reveal the identity of the person who created the message.
Unlike universities, schools often do not have an electronic database of students' personal records that can easily be interfaced with the user administration systems on the Internet servers. If no such interface can be created, a semi-manual system can serve as a temporary solution where teachers handle their students' requests, assign user names and passwords, and thereby provide the necessary user authentication.
Experience has shown that the connection of schools to the Internet poses particular security risks, even compared with the situation at universities where security problems always have been notorious.
Like university students, even teenaged pupils find access to Internet resources where so-called "exploit scripts" are published: pre-packaged computer programs which can be applied to certain server systems in order to gain incorrect privileges on these systems or to make the systems unusable for their rightful users.
But unlike the situation at universities, schools will usually not have their own security experts who track down misbehaving users and constantly monitor the security status of the server systems.
The necessity of securing computer systems and access to network services may sometimes even be seen as contravening educational goals: students shall be educated to cooperate and work as a group on a common task, but security restrictions demand that they each have their individual passwords and hide these from each other. It is their teachers' task to reconcile these conflicting requirements, in spite of the familiar atmosphere within a group or class of students.
Since most teachers will only have worked with their computer in a (presumably) safe environment such as their private home or a classroom, it will be difficult for them to get an idea of the threats to which an Internet user may be exposed. Sophisticated authorization procedures and secrecy about passwords, -phrases, and personal keys are acceptable only to those who understand why there is a need for reliable protection.
Educational networks for schools could be built as a separate infrastructure in parallel to the existing university networks. The cost for such a segregated wide-area network, however, would be disproportionate with regard to the amount of traffic that can be expected. On the other hand, the cost of networking over regular provider lines would be too high: a school that makes adequate use of Internet resources (e.g., presenting its own website, downloading audiovisual documents, conducting research over the Internet) generates an amount of traffic comparable to that of a medium-sized private enterprise, but the latter could easily afford a higher cost for the connectivity (particularly if that augments its business) which the school could not.
Since in many countries the universities are connected through a high-bandwidth backbone network, it should be possible to use free capacity within this network for other educational institutions without incurring much extra cost.
There are at least two popular ways to organize the backbone connectivity of educational networks and their interconnection to the Internet:
The "National Backbone" solution:
A fast (ATM or Gigabit) backbone extends over a large, inhomogeneous geographical area (like a whole country). It offers a number of connection points, mostly in the vicinity of one or several universities, an exchange point to private provider networks and to the major international and intercontinental links.
This model can be coupled with a "flat rate" cost distribution scheme where the connection cost for a university or a school depends only on the maximum bandwidth; this means that sites in areas with a well-established infrastructure and therefore (technically) cheap connectivity in effect subsidize those sites that need expensive links (e.g., in rural areas). In order not to disturb this bandwidth-based charging model, even smaller schools would have to pay their share of the total networking costs, or else the number of schools and other "small" sites must be strictly controlled so that the cost distribution scheme does not become unbalanced.
The "Regional Backbone" solution:
A fast backbone is operated only within a smaller area, maybe the realm of a regional political authority (e.g., a German "Bundesland"), and the educational institutions select one or more commercial backbone operators and Internet access providers who offer the required bandwidths at affordable prices. If the schools are represented through the local or regional authorities that supervise them, the number of participants to such a contract is still manageable; the fact that local providers are often cheaper than their nation-wide counterparts can contribute to a cost-effective solution.
This approach seems more flexible than that of the nation-wide network; it does not exclude smaller telecom and cable network providers, and it naturally adapts to the economical and urban structure of the region.
Connecting large numbers of schools to an educational network through normal phone lines or through individual leased lines is too expensive in all those countries where the traditional phone lines are still operated by a central telephone company that charges a high price on local calls. This cost is often not covered by a school's budget, and if the price depends on the amount of traffic or the duration of on-line connections, the cost cannot reliably calculated in advance.
There are several alternative solutions but their viability depends on the local situation. Many state-operated schools, for instance, have standard telephone lines to the phone exchange of the local city council, and if such lines can be used free of charge, they can be employed to establish ISDN or even xDSL connections to a central site from where a high-speed link to the educational backbone network is established at a comparatively low price.
Another solution, particularly for urban areas, is the installation of a wireless network that connects a school to a nearby university site or even to a private company that donates part of its own Internet bandwidth to tunnel the school's traffic to a suitable educational network node.
The technology for wireless networks as well as for the use of TV cables and the mains power supply lines for data transmission is currently rapidly evolving, and many practicable and cost-effective solutions are expected for the near future.
Few schools will be able to operate their own Internet servers (for the relevant services like e-mail, WWW proxy, Usenet news). Qualified personnel is needed for the setup and maintenance of these servers; system supervision, regular updates and backups, user administration and hardware maintenance require a considerable amount of time and experience. Only in very rare cases a teacher will be prepared (and sufficiently qualified) to perform all these tasks in his or her spare time. Therefore the servers will usually be concentrated at central sites (which also provide additional benefits like redundant hardware and uninterruptible power supplies). Depending on the technical infrastructure of the educational network, the servers can be deployed at computing centres of universities or city councils, or they can be operated by the Internet service provider.
Educational networks will always need direct or indirect financial support; they can only prove to be useful and convince students and teachers of the merits of the Internet if they are fast enough and provide sufficient capacities (in server space, number of possible users, etc.). To facilitate funding and a fair distribution of donations there should be some institution that helps with the acquisition and administration of monetary and physical assets.
Even if the network is purely commercially operated, the participants need some institutional way to articulate their desires, ensure an adequate quality of networking, and to organize common tasks (evaluation of traffic statistics; exchange of user information to grant users access to nearby servers of other organizations; establishing security procedures).
In the case of regionally organized networks such an institution can also be a party to the contracts with backbone operators and service providers. This may be in the form of a "charitable limited company" which is authorized to negotiate contracts and to accept and spend money for the operation of the network. In many countries such a company will be exempt from tax if it serves the single purpose of operating the educational network and not to shed benefits to any third party. It should be controlled by elected representatives from all the educational institutions which it is working for.
Operating the educational network through an existing authority (like a university or a local council) would also be possible, but often some of the participating institutions feel insufficiently represented by such an authority which they deem not to be impartial enough.
In these times of drastic cuts on public spending many countries cannot afford to finance network connectivity for schools to the same extent as they have done for universities; in order to provide schools with the necessary equipment as well as with the financial resources for network access new ways of funding have to be found. Public-private partnerships appear to be a solution to this problem which will become even more important in the future.