The Brave New World of Wireless Technologies: A Primer for Educators
Have you just received funding to install a network in one of your campus’s
older buildings? Great! Now what? Why not opt for wireless? Here, Jerry B'erner
offers some basic knowledge of wireless networking.
How do you retrofit an older building with today’s networking infrastructure?
If the building is built with thick interior walls, or with bricks, or is otherwise
architecturally unique, your options to run network cabling on the surface of
walls may be limited. You may also discover that you do not have easily accessed
spaces above the ceilings, (e.g., no “false” ceilings) into which
to install the cabling. And the installation of either copper wire or fiber
cabling may require thousands of feet of cable and many hours of painstaking
labor to connect. Are there any alternatives to the installation of physical
wiring
Five years ago the answer would have been a resounding no! Today we can safely
answer the same question with an emphatic yes. Wireless local area networks
(WLANs) are now feasible and available at reasonable prices from a variety of
vendors. However, the array of acronyms that face educators is imposing and
overwhelming; the technical staff in your networking group may not be fully
aware of the options. In most colleges and universities, networking personnel
are much more comfortable with wired local area networks (LANs). What is ‘Wi-Fi’
anyway? Do we need to understand the difference between FHSS (Frequency Hopping
Spread Spectrum) and DSSS (Direct Sequence Spread Spectrum)? While the technical
details of these implementations of wireless communications are beyond the scope
of an introductory article, we will examine the technologies available for implementing
wireless networks today and suggest guidelines for selecting the most appropriate
for our colleges and universities.
The Traditional Wired Network
Starting in the mid-1980s, the choice focused our attention to three aspects
of network infrastructure: cost, speed, and reliability–we could pick any
two! Therefore, we arrived at the following compromises:
- High speed and low cost: Ethernet at the sacrifice of relia bility
- Low cost and high reliability: AppleTalk at the sacrifice of speed
- High speed and high reliability: Token Ring at the sacrifice of cost
Today, this picture has changed substantially. But while our choices are much
different, we are still limited to two of the three elements when designing
a wired network. We introduce these concepts here so that as we examine wireless
technologies, we can focus on the same three factors–speed, cost and reliability.
The Wireless Network
A wireless LAN requires essentially the same components as a wired LAN: a network
interface card (NIC), a communication media (copper cable, fiber optics, etc.),
a set of rules that follow a standard (protocol) for interpreting the communication
signal, and an operating system that understands and maintains the connection
between the local and remote computer. What differentiates a wireless network
is the use of non-physical media for carrying our information between nodes
of the network. A wireless LAN uses several different bands of the frequency
spectrum; these frequencies are outside of the normal visual spectrum.
A wireless LAN adds to this mix a special device, the Access Point (AP) to
the network. This device communicates with both the wireless device through
the Radio Frequency (RF) signal and the wired network through standard protocols
like Ethernet. In addition, both the Access Point and the NIC must include antennas
to allow access to the RF spectrum. The key elements in designing a wireless
LAN are the number, placement, and configuration of Access Points for optimal
performance.
Wireless Networking Standards
Wireless LANs have been implemented in recent years via several standards.
However, only the IEEE standards defined by Project 802.11 have resulted in
sufficient production of devices that enable us to implement WLANs in our colleges
and universities. Two variations of this standard are now producing devices
for implementing our WLANs: IEEE 802.11b (11 Mbps in the 2.4 GHz band) and IEEE
802.11a (54 Mbps in the 5 GHz band). Both of these technologies function in
the unlicensed spectra: the former uses the ISM band while the latter uses the
U-NII band. Which one do we choose? Let’s take a closer look at the strengths
and weaknesses of these technologies.
IEEE 802.11b is the more mature of the two technologies, and devices (NICs
and Access Points) compliant to this specification began appearing in 2001.
These devices are available at a modest price and have an effective throughput
that is adequate for most normal WLAN applications in the office, lab, or classroom.
This throughput fluctuates as a function of the number of devices sharing the
channel, distance, and obstacles through which the signal must pass, but in
general ranges between 2 and 11 Mbps. The key to deploying this technology is
the placement of the Access Points and the assignment of computers to frequency
channels to spread the load.
The downside of the 802.11b standard is the relatively slow speed of transmission.
Since the maximum speed set by the standard is 11 Mbps, throughput for these
devices is the slowest of the 802.11 family of protocols. In addition, the 2.4
GHz range is relatively limited and has only three communication channels available–these
channels must be assigned carefully in areas of dense wireless use, such as
computer labs or in smart classrooms. Finally, the major weakness of 802.11b
lies in security; the use of the Spread Spectrum (DSSS) technology and Wireless
Equivalent Privacy (WEP) offers only very basic protection of the data being
exchanged.
IEEE 802.11a is the newer technology on the scene, with devices compliant to
this standard only appearing in the marketplace in 2002. This technology has
a higher throughput at short ranges (up to 10 feet), making it the candidate
for use in bandwidth-intensive applications such as multimedia. Unlike the three
channels available to 802.11b, 802.11a has eight separate communication channels.
This enables more computers to be connected without taxing the system. The use
of a multiplexing technology, OFDM, also provides for enhanced security, although
it still employs WEP as its major security component.
Again, on the downside, 802.
11a currently has several severe drawbacks. A major
disadvantage lies in the chip production technology required; chips cannot be
produced as CMOS chips, but must use more expensive technologies. Until the
volume of sales of these chipsets increases, 802.11a devices tend to be much
more expensive than 802.11b chipsets. In addition, it is a less mature technology
and must work through the refinement process. Finally, 802.11a is reliable within
a range of 10 feet, but the speed drops dramatically beyond 20 feet. This requires
that Access Points be located more closely to the users to maintain both reliability
and speed.
For a better understanding of wireless standards, we need to return to the
basic three elements that characterize a network: speed, cost, and reliability.
We must remember that we can have only two of these. So where do our wireless
technologies stand in this analysis? We have already stated that a designer
can choose which two of the three elements we value the most. When this analysis
is applied to wireless technologies, what do we get? Let’s take a closer
look.
- High speed and low cost: The IEEE 802.11a devices
- Low cost and high reliability: The IEEE 802.11b devices
- High speed and high reliability: At present, no wireless technology meets
these criteria; we must stay with our wired LAN technologies like fast, switched
Ethernet to meet this requirement!
Enhancements to the 802.11 standards offer solutions in the future, but our
present offerings are limited to those specified above. When products start
appearing for some of the emerging standards, this scenario needs to be revisited.
While a variety of technologies exist for implementing wireless local area
networks, the IEEE Project 802.11 family of standards is the most promising
at this time. Which of the 802.11 standards is superior? That question is harder
to answer, due to the differences in RF band (ISM vs. U-NII) and access technology
deployed (DSSS vs. OFDM). In the long run, 802.11g and/or 802.11i will probably
offer the most promising solution. These technologies are still awaiting implementation,
however. When faced with the choice of selecting 802.11a or 802.11b, the safest
choice will probably be 802.11b except in applications that require higher throughput
over a restricted range.
Security Issues
Security is the major drawback of wireless LANs. Basic security can be provided
by WEP. This is sufficient to start with, as long as the security features are
enabled and the NICs support 128-bit encryption. However, WEP is only basic
security and is not sufficient for critical data like student records or other
secure transactions.
Additional, sophisticated devices from Cisco Systems and 3Com are required
to allow more reliable authentication. These solutions are implemented by using
a combination of more advanced Access Points and a RADIUS server that maintains
a list of the authorized MAC addresses that are permitted to log into the WLAN.
Current Access Points allow the WLAN to be enabled without requiring that encryption
be turned on. In addition, many Access Points allow their Service Set Identifier
(SSID) code to be broadcast frequently, so outsiders may be able to gain access
to the WLAN. Proper configuration can provide a reasonable level of security,
but care must be taken to implement and use security options consistently.
The basic WEP protocol incorporated into most IEEE implementations is inadequate
for high-security applications. This problem can be resolved by selecting augmented
security schemas and authentication techniques. Future standards are being developed
to incorporate these changes into the wireless protocols. However, if high-security
applications must be accessed, the wired LAN is still the recommended.
Opting for Wireless
Wireless LANs should be seriously considered when upgrading campus networks.
The technology has matured and provides a positive set of advantages, including:
cross-vendor interoperability, practical interference-free communication over
reasonable distances, and minimal security that can be augmented by several
emerging technologies. Access Point devices have advanced to allow scalable
connection with the wired campus infrastructure, and students and faculty can
roam the campus and maintain their network connection. Above all, retrofitting
older buildings with networking using wireless technologies lowers the overall
cost of ownership and facilitates faster installation, simplicity, and flexibility;
and, except for connecting Access Points, physical wires no longer need to be
run through walls or between floors.
Characteristics of Wireless Networking on Campus
Mobility: The mobile user or student can move from classroom to classroom
with their WLAN-enabled computer and still access the Internet, file servers,
library resources, and so forth. This mobility can extend to conference
rooms or study rooms used by faculty or students in the course of their
planning and study.
Installation speed and simplicity: The use of wireless in older buildings
can save considerable money and challenges posed by retrofitting to support
networking. The wired LAN infrastructure can be minimized and used to
support the necessary VLAN connections and Access Points in the classrooms.
Installation flexibility: Flexibility is offered in both the networking
of dedicated computer labs and the use of mobile computer labs. This can
allow any classroom to become a computer lab, as needed, with the use
of computer carts. The primary requirement for a mobile lab situation
is a computer "drop" to which the access point on the mobile
cart can be attached.
Reduced cost of ownership: While mid- to high-end Access Points are not
inexpensive, the overall investment in the wireless infrastructure is,
in the long run, less expensive than retrofitting cables into old buildings.
And by not having fixed positions, rooms can be adapted for different
uses in the future without writing off the cost of the wiring; this is
especially true of computer labs where the number of drops would be substantial.
Scalability: The wireless LAN could start off small, perhaps with a mobile
computer lab, and then grow in size and complexity as needed and when
funds become available. Likewise, devices like Access Points can be upgraded
when the instructional needs and infrastructure dictate. The cost of current
Access Points can be migrated to other locations when new equipment is
purchased. This especially applies to computer labs.