The SSTA Research Centre grants permission to reproduce
up to three copies of each report for personal use.
Each copy must acknowledge the author and the SSTA Research
Centre as the source. A complete and
authorized copy of each report is available from
the SSTA Research Centre.
The opinions and recommendations expressed in this report
are those of the author and may not be in
agreement with SSTA officers or trustees, but are offered
as being worthy of consideration by those
responsible for making decisions.
This report focuses on "connectivity," that critical aspect of local and wide-area networking that makes effective computer linkage possible across the building or around the globe. Most of the guidelines for achieving effective connectivity will be found in Chapter VI. The earlier chapters simply develop some of the basic concepts for those who find this technical talk to be new and mysterious.
The networking of computers, both locally and over a distance, is creating a significantly more powerful environment for computer use than was generally available just a few years ago. The "intelligence center" of computer capability has shifted dramatically in recent years from the capacities and capabilities of a desk-top machine to what is now a network of machines and resources. These networks, whether confined locally within a building site, or comprising a much wider connection to machines and resources at greater distance, now collectively provide an information accessing and processing capability that is exceptionally powerful and useful. What has made this shift possible are the technologies relating to "connectivity."
Connectivity for schools are basically of two kinds: internal and external networks. The primary focus of this report will be on the need to effectively design and implement connectivity technology within the school. But one of the primary purposes of the local area network will be to connect effectively to one or more wide-area networks. Since the design and maintenance of networking beyond the walls of the school are typically the responsibility of major commercial telecommunications providers, this aspect of connectivity is treated only briefly in this report (see Chapter V).
Connectivity within the school building, involves significant questions
concerning wiring and the various components that allow computers to communicate
with each other and the outside world. The communication capability
of a network is dependent upon the quality and capabilities of each linked
component within this network. Further, the intended use of computers
within a building and their physical distribution within the site are all
factors that must be considered in designing a network that will function
efficiently. But experience demonstrates that networks are not static
entities, they are rather highly dynamic phenomena. Networks will
be expanded. Various components will be upgraded from time to time.
The concern for expandability and ease of maintenance must be considered
from the start.
While efforts have been made in this report to provide an overview
of current best practices, readers should be aware that communications
technologies are in a state of continuing development. Certain guidelines
outlined in this document are sure to be obsolete within coming months.
Nevertheless, the information provided here is intended to equip local
decision-makers with enough knowledge to enable them to engage in dialogue
with technical professionals who, normally, will be contracted to supply
and install networks.
A number of documents, mostly available on-line, are referenced throughout
this report. Indeed, their quality and availability are such that
little reason exists to duplicate the technical details here.
The report has been written for school board personnel who are faced
with major decisions on implementing and supporting computer-based communications
technologies. It attempts to address frequently asked questions and
lessons learned by school personnel who have pioneered development of school-based
networks.
The report addresses only one part of the larger picture that must be given attention by local decision-makers. Central to all decision-making should be the development of a system-wide technology plan. The development and maintenance of computer networks in schools are, and will continue to be, a significant budget item. It is essential that the educational purposes for using computer-based technologies are clearly defined and that all aspects of a technology-enhanced instructional system are developed in balance. Computer networks, while providing a critical element in a well-designed learning system, remain only one of a number of key components or tools. The technology plan should address these broader issues with a view to ensuring that a particular networking configuration is developed to support the intended instructional and administrative uses which are of the essential reasons for everything that happens within the school.
Connectivity can be defined in terms of how a desk-top computer can be connected by telecommunications links to other computers and information sources across the room, across the building, or almost anywhere in the world. Such connectivity makes possible the range of functions now commonly taken for granted: fax, e-mail, voice-messaging systems, online banking, videoconferencing systems, and a host of information sources and resources made possible by Internet access.
Connectivity for schools and school systems offer a number of advantages of current significance and growing importance:
Emerging patterns for computer use in schools
Clarifying the intended use of computers in the school must be viewed as a central or basic consideration from which other planning and design considerations should follow. Once we know what we want to do with computers, we can then define with more precision the type of equipment to be acquired, the physical location for distributing such around a school, and can develop strategies for effective use and ongoing expansion of facilities. Unfortunately, in many cases, school personnel do not always have a clear idea of precisely how equipment can and should be used. This is a new technology, and opportunities to become well acquainted professionally with the utility and power of its use in teaching and learning is a matter of on-going development.
Effective use of computers is integrally related to the availability of quality software. In contrast to the business community, there exists no well-defined set of software tools that are in any sense an "industry standard" in education. Comparatively speaking, schools are faced with complex decisions on the selection of software resources for different instructional levels and within various curriculum areas. The emerging nature of this technology is reflected in a myriad of software products, many of dubious quality and design of a confusing assortment of possible patterns of use with learners.
The development of a Technology Plan, as discussed more fully in Chapter III, is part of the process to systematically address what is possible and what can be considered reasonable for implementation by a particular jurisdiction within various time frames.
It exists both as a focus for learning about technology as well as learning with the new computer-based technologies. The former focuses on developing competence in the use of technology as a tool. The latter focuses on ways in which technology may facilitate the already existing broad range of required learnings (the curriculum) in schools. And within that framework lies a hundred potential uses for the technology, each providing a choice of what is to be learned and choices on how it is to be learned. In contrast to many business applications that become standardized about a few highly repetitive transactions, the child working his or her way through the curriculum seldom performs the same activity more than a very few times.
The acceptance of technology into the culture of a school is something like the proverbial chicken and egg. Which comes first, the technological infrastructure or the redesigned teaching/learning methodologies? Until facilities are in place, many teachers will not have opportunity to become acquainted with the emerging pedagogical practices which can use the technology effectively. On the other hand, until the advantages of using technology-enhanced teaching methods are understood and sought-after, technology in and of itself may simply lie unused. Clearly, both the infrastructure and its effective use must develop in tandem. The one reinforces the other.
Where do we locate computers within schools?
At a most basic level, computer use in schools has occurred primarily in one or more of three common locations: a computer lab, as part of the library facilities, or within classrooms. As machines are being added to schools, and as networks are becoming more common, other sites for instructional uses of computer use can be added, such as meeting rooms, staff rooms, other more specialized office areas and most anywhere that students and/or teachers can find some space to work individually or in groups.
Computer labs are appropriate for instruction on computer use with a class as whole. This may include instruction on keyboarding, word processing, and other applications. When basic computer skills are the object of instruction, teachers will most often prefer to have the whole class in a room where computers are available to students on a one-to-one basis. This use of computers has been the primary focus of Computer Applications, Computer Science, and Business Education courses offered in the past at the Junior and Senior High School levels. At the senior levels, therefore, most schools continue to find a significant need for one or more centralized computer labs.
For some time now, students at elementary levels have been expected to learn keyboarding and use of word processors. Currently, students at these levels are also being encouraged to use e-mail and the Internet. In some cases, schools will encourage the widespread use of multimedia programs such as HyperStudio which involves development of skills in scanning photos and developing navigable on-line collections of screen displays. In other cases, schools may be adopting some of the newer computer-enhanced approaches for teaching in curricular areas such as Mathematics. Some of these learning activities and approaches require students to be matched with personal computers for extended periods of time each week. In still other scenarios, only a few computers are required for students in a classroom.
The size of the school and the types of computer experiences planned for students will be the important factors in determining the need for one or more separate computer labs. While labs often provide the most efficient means for supervising the learning of intensive computer-related skills, they are often not the setting of choice for many other computer uses.
Disadvantages of computer labs arise from the shared nature of such facilities. Classes are often commonly timetabled into labs for specific times each week. While this may be appropriate for on-going programs which require intensive computer use throughout a term, it is often found inadequate for many of the periodic and shorter-term uses of computers that would fit more naturally into on-going classroom-based curriculum learning.
Computers in the classroom are the setting of choice for many of the newer uses of computers in schools. Advantages include the following:
The Planning Process
Decisions to engage in the development or expansion of both internal
networks and connections to external networks need to be considered within
an overall plan for technology use in the school's program. Providing
connectivity may be a significant cost to a school or school system, but
this cost will be justified only if all other components are developed
and implemented too.
Sources for information on development of technology plans exist on
the Internet. The following are three suggested sites:
Planning for Acceptance and Usability. Stakeholders need to be identified and consulted, from the provincial ministry of education, to potential suppliers of network components and technical support, to the parents of children in the school. Teachers, generally, and the specialized needs of various educational programs must all be taken into account.
Planning for Implementation. A technology plan will outline a step by step progression for not only putting the hardware in place, but for ensuring that the professional and technical skill development of staff is adequately taken into account.
Planning for Maintenance. Ensuring that the system is dependable is critical for its acceptance and use by all users. Technology has moved quickly from a type of "add-on" to existing practices, to an integral part of new approaches to achieving the educational functions of schools. It cannot be allowed to breakdown any more than we would plan to operate many schools with no buses or no electricity.
Planning for User Support. The number of software products anticipated for use in schools is many times greater than that used in most business settings. Teachers are also heavily committed to a host of tasks and interactions with students. Finding time to learn new procedures with technology is normally a problem. Initial training on new software products is only a beginning. Continuing support, by phone and in person, is critical to teacher acceptance and effective use of technology.
Planning for Scalability and Acceptance of System Limits. The implementation of a given network will force decisions that balance out the capabilities with the budgetary resources to implement and support the system. An initial system must have operational and support integrity. That is, it should be designed to function well within the confined capabilities chosen initially. With adequate long-term budgetary planning, the initial capabilities can be expanded, and various components added or replaced to achieve this growth.
For instance, a network designed for basic browsing of the Internet can be made to function well if a dozen student are expected to be using a common browser for common tasks at one time. But, there are constantly new Internet-based capabilities that will not function in this environment. Streaming audio or video is a case in point. Such uses of the Internet promise to grow in popularity, but simultaneous use by one or more students in some settings may severely degrade the network performance for other users.
Part of the design process, therefore, must anticipate the limits which are reasonable at any given time. Often the techno-hype associated with such developments tends to oversell the capabilities.
But planning for scalability anticipates the continuing need for expansion and improvement. Machines purchased today will have known capabilities. Machines purchased in three years will probably not cost any more, but will have considerably greater capabilities and speed. Budgeting should anticipate the need for such replacements. Planning the infrastructure should anticipate the need for excess capacity for those components that will be difficult and costly to replace in the future.
Planning for the acquisition of network components
This report does not attempt to provide even a minimal guide to the practices of technology purchase. Nevertheless, it may be important to note some of the common pitfalls and better practices that are known to be used from time to time.
With technology changing as quickly as it does, and with the complexity of technical knowledge and decision-making to design, acquire, and implement effective computer systems, it is important to recognize that the purchase of networked systems is far from a cut and dried procedure.
Few school systems have the technical expertise internally to fully understand the field, especially in terms of the continually evolving capabilities of many networking components. If a tendering process is chosen as the means for acquiring equipment, vendors will normally treat the defined needs simply in terms of costs and delivery rather than in terms of the effective integration of these products into a new or existing networking system. This approach, then, is inadequate in most cases. Components must be selected with care. Their effective integration and functioning within the overall system must be assured. Only if such considerations are given top priority will technology budgets be spent most effectively.
The need to create new systems or expand existing systems is often met through a process by which a commercial vendor is selected as a type of partner in the ongoing process of network development and use. The selection of such a vendor may be part of a request for proposals (RFP), in which the needs of the school or school system are made explicit and vendors are invited to submit their proposed solutions. While price of product may be a significant factor in decision-making, the existence and quality of professional support at every stage of planning, purchasing, deployment, and on-going support is possibly even more significant.
School systems often define the dollars they have available for development or expansion of their systems, then invite one or more vendors with whom they have created an on-going relationship to assist them in developing the system to meet their needs. Vendors chosen on the basis of their reputation and knowledge of the field will then work with the school system to cooperatively develop, possibly over a number of years a reliable functioning system that meets the needs of the school or system.
In summary, then, the environment for acquiring computer systems and networks is significantly different from the procedures often used when purchasing discrete items whose function is not so critically tied to an elaborate and sophisticated system.
What can local decision-makers do?
The more informed school system people become about networks, the better position they will be in to engage vendors on design matters and enter into contracts that will yield a quality system.
The following guidelines are critical when engaging in discussions with potential vendors.
Local Area Networks -- Basic Concepts
This chapter outlines conceptually the basic components of networks and their functions. In Chapter V we examine wide-area networks in a similar way. In Chapter VI more technical details and standards are discussed as guidelines to the installation of such components in a school site.
Networks, by definition, are comprised of links and nodes. A connectivity focus is primarily a focus on the cabling infrastructure, or equivalent, that links computers and other components within a network.
A local-area network or LAN connects computers and other shared peripherals such as printers and CD-ROMs in the same building or, at times, across buildings in close physical proximity.
A wide-area network or WAN spans a considerable distance, from linking schools in different communities to spanning the globe, as with the Internet. Indeed, most schools are primarily interested in Internet access, whether for such limited purposes as e-mail, or for access to the rapidly expanding resources residing on the World Wide Web.
Computers can be linked through both wired and wireless technology. The common types of hard-wire connections are twisted pair, coaxial cable (both of which use copper wires), and fiber-optic cable (made of tiny glass strands). Of these, twisted pair technology has gained ascendancy. Some wireless technologies (microwave) have been in use for decades. Current applications of wireless using spread-spectrum and laser technology are currently under rapid development.
Enhanced twisted pair is a term used to distinguish the current standard networking cable from that normally used with telephones. It differs in terms of size and quality, but uses modular block connectors similar to that used with standard interior telephone cable.
Fiber-optic cable is commonly used primarily over longer distances. Light can carry a digital signal much farther between nodes in fiber than electricity can carry a signal in twisted-pair or coaxial cable. It is often the cabling choice for connecting school buildings in close proximity on an extended LAN, and within buildings to connect wiring closets in larger installations.
Coaxial cable, one form of which is commonly used with cable TV, was once thought of as the logical emerging choice for high-capacity data transmission. But with the new electronics designed around the much more common twisted pair cabling, the greater physical dimensions, weight and cost of coaxial cable is not required for even very high-speed data communications. For this reason, coaxial cable is typically not used for LAN cabling.
Wireless forms of communications are most commonly used over longer distances and require line-of-sight transmission corridors between transmitters and receivers. Of these, microwave technology has been with us for years as part of the microwave grid used for telephone and TV communications. More recently, wireless is used to receive satellite signals, which allows even very remote communities and schools to achieve some level of service that would be impossible with land-based cabling systems.
Recent developments with wireless technologies are realizing improved bandwidth. For shorter distances up to 5 km, 30 Mbps is now possible. For distances approaching 50 km, signal transmission may approach 11Mbps under ideal conditions.
Under some conditions, particularly in older schools where costs of running cables proves to be prohibitively expensive, wireless LANs are being installed. Wireless in these settings, uses a type of spread-spectrum technology similar to that in use with cordless telephones. Using a system of transmitters and receivers, a maximum bandwidth of about 10 Mbps can be achieved within a localized area up to about 50 metres. Costs are high for this type of network connectivity, and maximum bandwidths are far below the 100 Mbps speeds of twisted-pair LAN technologies.
Listed below are highly recommended web sites where more in-depth information is available on the rapid changes occurring in wireless communications. Other product-related sites are listed in the References.
A node is any device that is connected to a network such as a computer or printer. One common type is a client device that requests and uses resources available from other computers or devices. When client nodes exist, there must also exist one or more server computers or devices within the network. (The terms 'client' and 'server' can the thought of in terms of the functioning of a restaurant, where clients make requests of servers who deliver to order.) Most often, the term 'server' will be used to describe a higher capacity computer having extra disk storage space and speed for sharing resources with client nodes.
At other times, however, the term 'server' will be used to describe other specialized devices such a print server (that manages print jobs for a network printer), a communications server (that manages e-mail), or a Web server that efficiently manages the requests for Web pages to and from the Internet. In other cases, servers are used as gateways that manage communications between a LAN and other LANs or WANs.
The term network architecture is used to describe how the components of a network are arranged and how network resources are coordinated. Part of this is something called the network strategy of which the two most common types used in LANs are the client/server system and the peer-to-peer system.
A client/server network system commonly uses one powerful computer to coordinate and supply services to all other nodes on the network. The server provides access to centralized resources such as databases, application software, and hardware.
A network operating system (NOS) is the software used to control and coordinate communications and data transfer between computers on a network. The two most common products providing this service for client/server network systems are Novell's NetWare, and Microsoft's Windows NT.
Client/server network systems have significant power and sophistication
for managing large networks efficiently. Such advantages, however, are
reflected in higher installation and maintenance costs.
A peer-to-peer network system is software that enables computers
on a network to act as both servers and clients. One computer can
obtain files located on another computer and can also provide files to
other microcomputers. Operating systems using this strategy include
Novell's Netware Lite, Microsoft's Windows 95/98/NT, and Apple's Macintosh
Peer-to-Peer LANs.
For relatively small networks, a peer-to-peer networking strategy has advantages in terms of ease of installation and lower cost. Their disadvantages include the lack of powerful management software to effectively monitor a network's activities.
The cabling infrastructure developed for one of these networking strategies will serve the other type of strategy well. In discussions below, a client/server strategy will normally be assumed but this is not intended to define which strategy may be most appropriate in a given situation.
A network gateway, or Internet gateway, is a special type of node that allows a LAN to be connected to another LAN or a WAN. Gateway technology can serve a number of specialized functions including firewall protection (preventing unauthorized external access), filtering systems (preventing internal users access to objectionable material or a list of prohibited sites), caching (the storing of frequently used Internet files for rapid local access).
In designing a school LAN, it is necessary to identify the different types of communication links and devices and understand their functions. In the analysis that follows, a client-server type of network is assumed. Central to the functioning of the network will be the existence of the server. Connections from the server to the various computers, which might number dozens or hundreds in a school, are managed in a configuration commonly known as a "star formation" through a series of switches and hubs. That is, we can view this setup much like a wheel with spokes with the file server connection to a switch or hub at the center and workstations at the peripheral points. Some spokes will be more complex than others, with hubs or switches facilitating a large number of computers to be connected to a single centralized access line to the fileserver.
Cables normally terminate as RJ-45 connectors in a wall-mounted outlet. Classrooms equipped as computer labs can be expected to have 30 or more such drops located at appropriate points of convenience for the expected arrangement of computers within the lab. Classrooms will be expected to have possibly a half-dozen drops, libraries perhaps 8 or 10 drops. Offices, larger meeting rooms, the auditorium, etc., all need to be assessed for possible future use of computers, and a suitable number of drops located for convenient access by users.
Although not strictly part of the LAN, the network design must also anticipate the need for power outlets at all locations where computers may be connected to the LAN within the school.
Within the wiring closets are located one or more patch panels which provide a flexible and convenient location for connecting cables to hubs, switches, and servers. All of this equipment functions best with a reliable and clean power supply provided by a UPS (uninterruptible power supply). Racks or shelving units are required in the wiring closet to hold all of the electronic equipment. Often located in the wiring closet, too, is the terminus for the outside WAN connections, and possibly even equipment supporting the telephone system within the school.
If the primary server for the network is not located in the wiring closet, it will use one of the many cable links into the wiring closet to achieve connectivity to the rest of the network.
The wiring closet functions not only as a collecting point for cables from the various locations in the building, but also as a convenient location for managing the various cabling connection necessary to achieve appropriate balance over the network.
Locations of wiring closets are determined by the lengths of cables required to serve the school premises. Twisted-pair cabling has a standard maximum length of close to 100 metres over which network devices can reliably communicate. The closet must therefore be strategically located within a building to ensure that cable runs do not exceed maximum lengths specified by the standards. If one location cannot serve all points, then auxiliary closets must exist.
The collection of equipment in the wiring closet provides a one-stop maintenance location for the physical aspects of connecting and managing the network. One or more computers may be housed here functioning as either the primary, or other specialized servers. And since all of this electrical equipment generates heat, a wiring closet must be appropriately ventilated. In larger installations, this means air conditioning, since heat dissipation is not just a concern of human physical comfort, but is a critical matter for preventing electronic devices from becoming overheated.
What is meant by "the electronics"
Communications over cables always requires some type of electronic device at either end. These devices allow "packets" of digital signals to be sent and received. These devices are chosen according to the type of communication signal to be driven over the cable and the capacity for transmitting data as bits of digital information per second (bps). (The term bit in this case, is the electrical equivalent of a 0 or 1 and should not be confused with the common term byte which is a standard sized collection of bits (commonly 8, 16 or 32 bits) which describes a larger unit of information, such as a letter or a number.)
Ethernet is an industry standard LAN protocol for routing digital packets of information from source to destination nodes throughout a network. All computers and electronic devices attached to such a network must be ethernet-compatible in order to communicate across the network. Computers need to be appropriately equipped with internal ethernet-compatible network interface cards (or NICs). Most Macintosh equipment and some PCs are delivered with an “on-board” NIC. All other computers must have an ethernet NIC installed in order to be connected to the network.
Hubs and switches are common components for most school LANs. These devices provide a one-to-many type of connection, often with 12, 16 or 24 connecting ports. Hubs lack intelligence that switches have. Switches have the built-in “electronic intelligence” which allows them to be configured and managed remotely to maximize network traffic flow.
To understand how an ethernet network functions, it is important to understand that a communication stream of data (consisting of on/off bits) is broken up into packets of data, each of which is sent out over the network and for which confirmation of its delivery must be received in one form or another by the sending device. If a packet is not received as intended, the sending device will resend the packet.
The simplicity of the ethernet standard for such communications is based on the real possibility of interference in signals when more than one device chooses to send a signal simultaneously. When this happens, we say that a collision has occurred. Ethernet communication protocols (the distinctive patterns by which communication is achieved) enable an ethernet device to "listen" for existing traffic on the network so as to time the release of a digital packet onto the network. If this happens to occur, however, at the same time as another device is sending data, the resulting collision can be detected and each of the devices will attempt to transmit their signals again after a randomly defined time delay.
The potential for overloading the LAN, or some portion of the LAN, by one or more intense users, leads to design decisions which can isolate heavy usage in one area to protect priority usage in some other part of the network. Switches have the capacity to direct and limit the transmission of data packets according to the addressing information contained within the packets. Most hubs have no such capability and function only to provide hardwire connections of many cables to a single cable for purposes of bringing signals to a common destination. The process of adjusting the cabling infrastructure through selective use of switches, hubs and the cables themselves is a process known as "balancing" the network. A well-balanced network will provide satisfactory performance for all users, even if some users have more intense needs for LAN use than others, or will isolate lower priority users such that priority users will be generally protected from excessive use by others.
To illustrate further the load limits of a LAN, let's consider a scenario where thirty students in a computer lab are asked to access a Microsoft Word file located on the fileserver. This is not normally considered to be a heavy demand on a typical network. However, if 30 students choose to download some large graphic files simultaneously to machines in a lab from the fileserver, serious degradation in performance is likely to occur for other users on the network until this heavy demand passes. Advances in audio and video streaming (where a continuous flow of audio or video signals occurs to the workstation) can create exceptionally high demands on a network. Even one or two users of this type of technology can cause a serious reduction in performance on lower-bandwidth systems for other users.
In situations where administrative functions are performed on a network used by students, switches should be appropriately located within the system to ensure priority use is available for priority users.
The ethernet protocol for managing data packet communications on the network should not be confused with the network operating systems such as Novell or Microsoft NT. It is common to talk of "layers" within the design and functions of a network. The electronic layer that control the transfer of digital packets exist as a separate layer from the network operating system which exists as software on the server and other client computers of the network.
Ethernet, as the leading networking standard, functions well with all
of the major brand-names of computers and common network operating systems.
Network interface cards (NICs) are now commonly configured to support the
multiple protocols which have evolved with different systems. Some
common protocols are:
IPX/SPX | commonly used by networks having servers running the Novell operating system |
TCP/IP | the protocol required for Internet connection, and now becoming a standard for all LANs |
NetBEUI | common to systems using Microsoft operating systems (95, 98, NT) |
AppleTalk | common to Macintosh systems. |
Recommendations from bodies such as the SchoolNet Advisory Board call for increased standardization on the TCP/IP protocols for all educational applications. This would have the effect of making translations of protocols unnecessary within and across networks, thus further increasing the speed of transfer.
One of the major decisions in designing a network is to select an appropriate capacity and speed for network communications. Once this decision is made, all network components must be selected to provide, at a minimum, the desired bandwidth standard.
The term "speed" is often a confusing term. In one sense, all networks pass data at the speed of light. Yet, if one has massive data to transmit, some connections will pass this data much more quickly than others.
The common units of measurement for describing the ability of a communication
link to transmit data are summarized in Table 4A.
|
Unit | Speed |
Bps | bits per second |
Kbps | thousand bits per second |
Mbps | million bits per second |
Gbps | billion bits (or gigabits) per second |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
As high capacity users flock to the new standards, prices will drop and we can anticipate that 100 Mbps equipment will be less expensive in a few years than is 10 Mbps equipment now. In the not too distant future we can anticipate, as well, the demand for 1000 Mbps (or 1Gbps) capacity.
The electronic components must be able to communicate with each other at whatever communication standard is selected. Some devices, such as switchable hubs, are dual mode devices in the sense that, currently, they are able to detect communications at either 10 Mbps or at 100 Mbps and adjust their functions accordingly. Such devices are connected to fileservers functioning at 100 Mbps and computers and other electronics operating down line at the slower 10 Mbps speed. However, many workstations are also now being connected at the 100 Mbps standard.
Confirming a network's functional capacity
As a network is being assembled, it is critical to ensure that every cable and every component added to the network is free of defects that would prevent it from performing at the critical minimal level. There are testing instruments and industry-wide standards which are used to "verify" or "certify" that all components and segments of the network are capable of performing to specifications.
Defective components as well as shoddy assembly of connectors and other components can effectively destroy the communication capacity of what otherwise is an expensive and high-performance networking system. Suitably qualified and/or certified technicians can be expected to deliver a superior product which can be shown to function to expectation through verification and certification procedures.
Industry standard specifications have evolved over time to address the continually increasing demand for higher speed and reliability. Networking standards have been outlined from time to time by professional associations to ensure that a certain type cabling and its related electronics will perform with reliability at its intended speeds. The current specifications of interest to school clients is outlined in the TIA/EIA-568-A Commercial Building Telecommunications Cabling Standard, October 1995. This standard is based on the use of cabling defined as Category 5 or 5e which has the capacity to transmit data at 100 Mbps. These standards are of sufficient stringency that specialized training and the use of specialized tools for installation and testing are now required.
New schools should now all be adequately designed to facilitate the running of cables and the housing of electronics and network components in wiring cabinets. Retrofitting schools that lack these amenities is often costly.
Wide-Area Networks -- Basic Concepts
Wide-area network (WAN) technology deals with connectivity over long distances, normally from 1 km to thousands of kilometres. The idea of a metropolitan-area network (MAN), or the type of network designed within a school division to connect schools at distant sites, will be considered below, also, as a WAN. A WAN is a form of external connectivity.
For many users, access to wide area networking is achieved through a local telephone connection. Since telephones use analog or sine-wave type technology, a modem is necessary to convert the digital signal used by computers and local area networks to the analog signal common to telephone lines. To gain access to the Internet by this means, users must also engage the services of an Internet Service Provider (ISP) which provides the on-ramp to the major wide-area communications channels.
There are a number of other higher-speed communication links to the Internet and other forms of wide-area networks. Whichever type of link is chosen, a gateway of some type is normally required that allows the LAN to link to the WAN connections.
WAN services are most often brought to a school building by a commercial telecommunications carrier. This is most often a telephone company (sometimes referred to as a "telco") or a cable company. At other times, a school division may provide local WAN connections using line-of-site wireless technology. Depending upon the geographical location, schools may have choices of carriers and choices of services of different bandwidths. In smaller population centers (outside the nine major population centres in Saskatchewan), little choice of commercial carriers and types of services may exist. In remote locations, no service beyond standard telephone connection may be available except as possibly provided through the installation of an expensive satellite connection or anticipated new wireless services being developed.
Providing connectivity to the Internet is commonly accomplished in one of the following ways:
Linking building sites as part of
a LAN
Constructing a LAN across different buildings in close proximity (within
a town) is commonly achieved in one of the following ways:
The bandwidth needs of a school must be realistically assessed by taking into account the following factors:
The ratio of active access and passive viewing makes a significant difference in the number of active users who can be accommodated on a WAN. When Web sites consist primarily of text, then many more users can be accommodated simultaneously without serious performance degradation. If the Internet access consists of multimedia with much graphical information and audio/video clips, then active downloading will be a high proportion of a user's time on-line and many fewer users can be accommodated with the available bandwidth.
The ratios used to derive expected needs for bandwidth to a school can be calculated by assessing the extent and type of use expected within a particular school. One heavy multimedia user may tax a WAN connection to its limits. At other times, 10 users might achieve acceptable performance with the same WAN connection if they are primarily accessing text and are engaged in reading of the downloaded files.
Too many externally connected users for the available external bandwidth will result in a quality of service issue. If students need to wait inordinately long, say 20 seconds, for a web paged to be downloaded, one can count on frustration being experienced by both students and teachers.
Connectivity speed "to the desktop"
The sharing of external connectivity among all active Internet users has the effect of distributing bandwidth among the users. If the WAN is capable of delivering 1.5 Mbps and there are 50 active simultaneous users, then each receives the equivalent of only 30 Kbps to their desktops. This performance is slightly less than what might be expected for a single user using a standard 56K telephone connection (for which 33.6 is normally considered to be the effective rate). Conversely, if it is determined that a single user requires 33.6 minimum bandwidth, then more than 50 simultaneous users can overload a 1.5 Mbps external connection.
Since connectivity speed to the desktop is the effective speed of the Internet for the dedicated use of a single user, the calculation of bandwidth requirements for a school is a complex matter. And if, as is sometimes the case, a number of schools share one Internet connection, the calculations must take into account the expected simultaneous use across all such schools.
Bandwidth requirements by type of Internet access
The Industry Canada funded SchoolNet National Advisory Board (SNAB) uses three classes of usage relating to the type of learning applications a school might choose to access. The full document is available at
|
Basic Web browsing/use | Supports basic Internet applications such as web-browsing, file downloading and Usenet access; low-end audio and video applications; e-mail and real-time text conferencing tools such as ICQ. | Requires 33.6 Kbps to the desktop2 |
Limited multimedia use | Supports more advanced Internet applications using more intensive use of graphics, audio and video. Applications such as Microsoft's NetMeeeting is supported with audio-conferencing and shared white board facility. | Requires 128 Kbps to the desktop2 |
Full two-way multimedia | Supports applications such as video-conferencing simultaneously with limited multimedia-based Internet access elsewhere within the school. | Requires from 384 to 512 Kbps to the desktop.2 |
The scenarios outlined in Table 5B may give some idea of the performance
associated with certain bandwidths. The table reflect experiences
of a number of specific Saskatchewan schools.
|
|
|
|
|
|
1 | 56K dial-up modem | Internet router software e.g., IPRoute or Wingate | Basic Web browsing | 5 or 6 |
2 | Dedicated 56K line | Caching web server | Basic Web browsing | 25 |
3 | 256K dedicated line | Caching web server | Basic Web browsing with one or two users accessing sites requiring high bandwidth | 30-40 |
4 | T1 connection | Caching web server | Basic Web browsing with a few users accessing sites requiring high bandwidth | 120-150 |
5 | Satellite downlink with 56K dial-up | Limited caching | Basic Web browsing with one or two users accessing sites requiring high bandwidth | 20-25 |
Current goals recommended for bandwidth to schools
As a short-term goal (to March 2001), SNAB recommends moving all schools towards the Limited Multimedia class of external connectivity. (See Table 5A above.) In practical terms, however, estimates of actual usage should be developed from anticipated usage patterns at different grade levels and actual program requirements. An argument can be made, for instance, that primary level children often use multimedia resources to a greater degree than secondary level students who function well with text. For this reason, the estimation of required bandwidth might be better served by considering the requirements of different segments of the overall school population.
Techniques for extending effective bandwidth
A number of techniques can be used to improve performance access to WAN connections. The following tips are suggested by Alberta’s Network Design document (p 59):
The physical connection to the WAN
Having estimated the need for bandwidth for a school, and having considered the costs, an agreement will be made to bring the external Internet connection to the school site. The demarcation point where the LAN connects to the WAN is often in the wiring closet. If some other box is installed of more convenience to the external service provider, a connection is normally achieved between this location and the wiring closet using fiber-optic cable (if distances are more than 100 metres), otherwise UTP cable is adequate.
Depending upon the type of service provided by the telecommunications link, there is a need for some form of conversion from the type of communications protocol common on the LAN to that which is necessary for WAN communications. This may require a modem for slower 56K telephone lines, or other gateway devices for higher-speed digital connections.
LAN communications most often follow the ethernet protocol. WAN communications appropriate to Internet access uses what is known as the Internet Protocol (IP). The connection between the LAN and WAN networks is known as a gateway.
The term Internet ramp is frequently used to describe a variety of technologies used to achieve connection to the Internet. In urban settings, this is sometimes referred to as the "last mile" connection since the high-speed fiber-optic "backbone" system used by the telecommunications industry will pass through all urban centers. In rural and more remote locations, the cost of leasing lines or constructing alternative links can be a costly budget item.
Relative speeds of different Internet connections
Some of the typical technologies for gaining access to the Internet
are summarized below in Table 5C.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
A cable-based service is also asymmetric with theoretical incoming speeds of up to 30 Mbps and 10 Mbps outgoing. Commonly this bandwidth is shared by multiple subscribers resulting in an effective bandwidth to a particular school much closer to 1.5 Mbps.
ISDN is a type of circuit switched digital technology that uses dedicated connections from a school to an Internet Service Provider (ISP). Two 64 Kbps digital channels are combined to create a 128 Kbps bandwidth connection. Other dedicated connections provided by the telecommunications industry can provide various bandwidths of this type from 64 Kbps to 1.5 Mbps (defined as a T1 service) to 45 Mbps (T3).
Various types of wireless carrier services have been described above in Chapter IV. The reader may also be interested in the discussion of wireless technologies as provided by the Schoolnet National Advisory Board.
This section summarizes standards for cabling. For a fuller treatment of this topic the reader is referred to a document prepared by the Calgary Board of Education.
Existence and significance of industry standards
Organizations within the telecommunications industry specify cabling standards to which both manufacturers and installers are expected to adhere. Conforming to industry standards maximizes performance of a system, ensures compatibility with attached electronic devices, and delays obsolescence or the need to upgrade a system.
The TIA/EIA-568-A standard is currently used by industry when cabling
commercial buildings. It addresses standards for cabling and electronics
suitable for transfer rates from 10 Mbps to 100 Mbps using unshielded twisted
pair (UTP) copper cable. This standard is commonly known as the Category
5 or the revision known as Cat 5e. (At time of writing, the Cat 6
standard is being established, with work already underway on Cat 7.)
Each standard category or level anticipates higher bandwidth performance
for network cabling systems. Although the standards encompasses both
UTP cable and fiber-optic cable, practical considerations currently favors
UTP copper cabling for schools.
To achieve overall network performance up to expected bandwidth potential,
all components of the network, from network interface cards, cable connectors,
outlet hardware, switches and other electronic devices through to properly
pulled and protected cable itself must be installed according to industry-recognized
standards. Using substandard patch cords, for instance, in the wiring
closet can provide the weak link that can cause serious degradation of
performance over the whole system.
In designing the cabling network for a school, it is important to think ahead to possible future expansion beyond immediate needs. For this reason, virtually all areas of the school should be assessed for potential use of information technology devices, from classrooms to hallways, and from gymnasiums to administrative offices. The relative costs of labor and materials is such that installing more cables than may ever be needed is often just a minor addition to costs. Drawing fewer cables than needed can result in costly retrofitting in future years. When drawing cables to an area such as a staff room, for instance, it may be advisable to pull four cables rather than two. The costs will be only marginally higher, but the increase in potential use is significant. (The Calgary Board estimate is that the cost of running four cables to an outlet rather than two increases the specific cabling cost by only 20%.)
In all areas where computers and related devices are to be connected to the network, the following general guidelines apply:
Guidelines for wiring closets are as follows:
Planning and Budgeting
for Total Cost of Ownership
Total cost of ownership (TCO) is an important budgeting concept which
draws attention to costs beyond the specific costs of computers, network
components and basic installation costs. The TCO concept anticipates
ongoing costs relating to usage, support, and network expansion.
A number of critical decisions will be made in the development and implementation of any network. Questions will include the potential for expansion, the availability of local technical support expertise or the potential for securing remote on-line support services. Questions should also address the common temptations to be on the leading edge of technology change, as opposed to being content with more mature systems which are often available at lower costs and greater reliability. The appropriateness of some of these decisions can have a profound impact on the level of service provided to users, the reliability of the system put in place, and the cost of ownership of a system.
Unfortunately, there are no easy answers to some of these design questions. Small networks with relatively few computers can become large networks in settings experiencing rapid population growth. The selection of certain network devices, such as hubs or switches may serve one level of implementation well, but will not serve an expanded system adequately. One implementation may rely on relatively low-level technologies that require relatively less sophisticated knowledge to maintain. On the other hand, more sophisticated technical components may promise an improved level of performance but may be largely a waste of resources if trained support personnel are not available to provide required technical support.
The following guidelines are advanced to assist decision-makers when
developing their long-term technology plans.
Alberta Education. (February, 1999). Computer network security: Best practices for Alberta school jurisdictions.
Alberta Education. (February, 1999). Network design: Best practices for Alberta school jurisdictions.
Alberta Education. (No date). Technology implementation review: Grande Yellowhead Regional Division #35 Wolf Creek Regional Division #32. Best practices and key learnings with respect to technology, its implementation and management in education.
Calgary Board of Education. (February, 1997). CBE local area network environments: Standards and guidelines: Cabling. Version 2.0.
Fitzgerald, Sara. (June, 1999) Planning for the Total Cost of School Technology. A downloadable PowerPoint presentation.
Mississippi State University. (1996). Guidebook for developing an effective instructional technology plan. (A graduate student project, Dr. Larry King, instructor.)
National Center for Supercomputing Applications, University of Illinois at Urbana-Champaign. (1996). K-12 networking infrastructure guide recommendations. Published on the Web by North Central Regional Educational Laboratory.
Northwest Regional Educational Laboratory. (February, 1998). Guide to networking for K-12 schools. A project of the Washington State Office of the Superintendent of Public Instruction.
SchoolNet National Advisory Board (SNAB). (October, 1999). Report on connectivity.
Vicom Technology Inc. (2000). About This Wireless Networking Q&A. (A Web-site document.)
Wireless LAN Alliance. (No date.) Introduction to wireless LANs. A downloadable Web document.
Drake, Hugo. (September, 1999). Mix & match blues. Structured Cabling, vol 3, no 5, pp 12, 14.
Goyette, Line. (April, 2000). “Cat. 6: Does it exist or not?” Structured Cabling, vol 4, no 3, pp 28, 30.
Koperea, Paul. (April, 2000) Fibre-to-the-desk: The ultimate cabling system. Network World Canada, vol 10, no 8, pp 33, 37-38
Michelson, Marilyn. (February, 2000). A Question of Standards. Structured Cabling, vol 4, no 1, p 6.
Web sites of companies involved in wireless technologies
Cisco Networks
Back to: Technology