P1: C-167
Guizani WL040/Bidgolio-Vol I WL040-Sample.cls July 16, 2003 10:7 Char Count= 0
826 WIRELESSCOMMUNICATIONSAPPLICATIONSremained a niche, used only in environments such as
warehouses, where mobility is essential.Wireless LAN Standards
The main problem faced by early wireless LANs was that
there were no real standards. Each company produced its
own proprietarily systems. Many simply chose not to buy
a wireless LAN at all, for fear of being locked into a sin-
gle vendor’s technology. Standards began to emerge in the
late 1990s, but potential customers were again spoiled for
choice. Most wireless LAN standards are spawned by ei-
ther the IEEE or ETSI. Between them, they have produced
at least six incompatible standards.
Table 3 gives a summary of the wireless LAN standards
available.IEEE 802.11
IEEE 802.11 was the first wireless LAN standard to be
defined. It uses the same switching protocols as wired
Ethernet, but gave up wire in favor of ISM radio. Just to
make things more confusing, it is actually two standards
in one, each employing an incompatible type of spread
spectrum.
The simplest version of 802.11 uses a frequency-
hopping spread spectrum, rapidly cycling between fre-
quencies many times each second. It has 70 different fre-
quencies to choose from, so while some may be blocked,
another is clear. If information doesn’t get through, it is
resent. A more complex version uses a direct sequence
spread spectrum (DSSS). It transmits on all frequencies
simultaneously. This increases the data rate, but also uses
more powers.IEEE 802.11b
IEEE 802.11b is based on a DSSS version of IEEE 802.11.
It uses a better modulation technique to increase capacity
up to a maximum of 11 Mbps. As well as a large increase,
this speed also pushes to match the original Ethernet
standard.IEEE 802.11a
IEEE 802.11a reaches a speed of 54 Mbps by abandon-
ing ISM and spread spectrum. It uses the U-NII band and
a technique called coded orthogonal frequency division
multiplexing (OFDM), which is designed to minimize in-
terference caused by a signal reflecting off walls.HiperLan 1
HiperLan 1 was the standard proposed by ETSI in 1992,
when it first recognized the need for high-speed wireless
LANs. Much of it is based on Ethernet, although its radio
access technology was taken straight from GSM.HiperLan 2
HiperLan 2 uses the same spectrum as HiperLan 1,
but is otherwise much closer to IEEE 802.11a. It has a
maximum data rate of 54 Mbps, which is achieved by us-
ing the same coded OFDM technology.HomeRF
HomeRF is designed for home networking. It is based on
the original FHSS version of 802.11, and designed pri-
marily for price rather than speed. Its major innovation is
direct support for telephony. The other standards are all
designed for data only, needing additional software to be
able to carry voice.Access Points
Wireless LAN systems can all be used for ad hoc network-
ing between two or more users who happen to have a card
installed. Areas such as offices or homes can also be fit-
ted with access points, which both extend the range of
the system and enable it to link to ordinary LANs or the
Internet.
Private users can create their own miniature cellu-
lar network by setting up several access points. The en-
tire standard includes a handover mechanism, similar to
those of public cellular networks. The IEEE 802.11 fam-
ily inherits the soft handoff system from CDMA cellular,
meaning that the mobile unit tries to form a link with a
new access point before it disconnects from the previous
one. Hyper LAN systems use a hard handover similar to
GMS, which means that they must disconnect from one
access point before reconnecting to another. This is less
reliable and results in a short interruption in connectivity.
Access points can take two forms, hubs and switches.
A hub is the simplest, simply rebroadcasting everything it
receives. A switch is more discriminating, sending trans-
mission only to smaller subgroups known as segments.
They improve network performance, as the total capacity
is shared per segment.
On a wired network, there is no limit to the number
of segments. A switch could put every user on its own
segment, giving each access to the full capacity all the
time. Wireless networks cannot do this, because each seg-
ment needs its own spectrum. However, they could make
it much easier to reallocate machines between different
segments, a task that usually requires rewiring. A wire-
less network can do this at the push of a button, or even
automatically.WIRELESS ATM
ATM has been advocated as an important technology for
the wide area interconnection of heterogeneous networks.
In ATM networks, the data are divided into small, fixed
length units called cells. The cell is 53 bytes. Each cell
contains a 5-byte header that comprises identification,
control priority, and routing information. The rest of the
48 bytes are the actual data. ATM does not provide any er-
ror detection operations on the user payload inside the
cell, and also provides no retransmission services, and
only few operations are performed on the small header
(Guizani & Rayes, 1999).
There are some important factors that might help
ATM make it successfully in the wireless world. These in-
clude flexible bandwidth allocation and service-type selec-
tion for a range of applications, efficient multiplexing of
traffic from bursty data/multimedia sources, end-to-end