WLAN Equipment
This section describes various tools needed in a WLAN,
including the following:
Note
Before surveying, it is important to understand what the
customer applications will require for bandwidth and range. This may make a
difference in what technology will work best for the site.
You have a wide variety of WLAN products from which to choose.
The customer's network design should have the technology defined (or
recommended) as well as the product models. The survey engineer must verify that
these will, in fact, work for this site. The survey should not only provide the
location, antenna style, configuration, and so on, it should also actually
verify that the customer's design will result in adequate performance. If you
have any doubts, raise them with your customer (and document them in the site
survey report).
Access Points
Many users believe that for a survey, any access point (AP) will work and that they all work
similarly. Such a belief strays very far from the truth. As you learned in Chapter 5, "Selecting the WLAN
Architecture and Hardware," APs vary drastically with regard to technology,
style, capability, and performance. To complete a successful survey and
installation, you must use the AP that is intended to be installed. As with any
radio technology, the performance from one radio device to another can have
drastic differences. This is even true of different radio models from the same
manufacturer.
A comparison of two of the more popular APs in the enterprise
industry todaythe Oronoco AP2500 with an 802.11b radio and a Cisco AP1100 with
an 802.11b radioshows the variations that can occur between different products.
The Oronoco has a transmitter power of +15 dbm (30 mW), whereas the Cisco AP1100
has +20 dbm (100 mW). That is a 5-dB difference. The difference of transmitter
power between these two devices can make as much as a 30-percent variation in
cell coverage.
Another difference that occurs quite often is variation in
receiver performance. The Cisco AP1100 has a receiver sensitivity of 85 dBm when
operating at 11 Mbps. Yet the 802.11b radio in the Oronoco AP has a receiver
sensitivity of only 82 dBm. Although 3 dB may not seem like much, it can have a
dramatic effect on the fringe areas.
Figure 7-1 shows various
APs that you can use for site surveys. Notice the variations between antenna
styles, capabilities (attached versus remote), and physical size.
Looking at two APs from the same company, the Cisco AP1100 and
AP1200 both have the same radio and the same radio specifications. Because of
the antenna variations, however, they are miles apart (figuratively speaking) on
coverage performance. The AP1100 has an internal dipole antenna, and unlike the
AP1200 it has no options of adding external antennas and no way to improve
coverage performance by changing antennas. The same drawback is true of the
Avaya AP5 AP.
Another common issue that comes up when trying to select an AP
for a survey is the ability to use various antennas. Antennas must be certified
with the radio in many regulatory domains. Some vendors have a limited number of
available antennas. The Enterasys Roamabout R2 with an 802.11b radio, for
example, offers an external antenna capability, but does not enable you to use
external diversity antennas. In addition, the selection of certified antennas is
very limited.
Be aware that some of the centralized intelligence
architecture products require you to have the controller or switch along with
the AP; otherwise, the AP will not operate. This can be quite a cumbersome
solution for surveying.
Surveying with one model AP and installing another can create a
major problem in the system when it is installed. Bottom line, and it cannot be
stressed enough, survey with the same product that will be installed!
Client Devices
It is also of utmost importance to know and understand the
client that will be used in the site. Although it is not practical to survey
with every client that will be used, it should be noted in the survey report
what clients will be used, what the radio transmitter and receiver
specifications are, and what type of antenna is used on the device.
For example, on the pre-site survey report, you see that a
warehouse may use various clients, including a standard laptop, bar code
scanners, and Voice over IP (VoIP) phones. The
warehouse manager and a few of the employees who need network access in both the
warehouse and the office will use the laptop. The radio card used in this device
is the Cisco PCM350. These radios provide a transmitter power of 100 mW, and a
receiver sensitivity of 85 dBm when running 11 Mbps (802.11b). They also use a
built-in antenna that has a unity gain (0 dBm).
The bar code scanners may be mounted to the lift trucks. These
devices use the Symbol radio that has a receiver sensitivity of 84 dBm. Again,
the scanners have a transmitter power of 100 mW, and use a 5-dBi antenna mounted
to the roof of the vehicle.
Also in the warehouse is a Symbol Netvision 802.11b VoIP phone,
with an output power of 60 mW, a receiver sensitivity of 82 dBm, and an embedded
antenna that, as a best guess (it is not documented), has a gain of 2 dB or less
(based on size).
So which client do you choose to survey with? Unfortunately
there is no simple answer. However, you can make certain assumptions. First, the
device that has the minimum receiver sensitivity should be viewed. From this
list, the phone, with a sensitivity of 82 dBm, probably has the worst-case
receiver. It is down 3 dB from the Cisco client card. Because the phone antenna
is not exactly known, assume it is no more than 1 dB, giving the receiver a
total receive performance of 83 dBm. Still, it has the lowest performance of the
three devices cited on the presurvey report.
For the transmitter side of the clients, you also need to
consider the antenna as well as transmitter power. In this case, the lift-truck
system has a 5-dBi gain antenna, so it will by far have the best performance.
The PCMCIA card in the laptop has 100 mW with a 0-dBi antenna, and the phone has
a 60-mW radio with a minimal-gain antenna. So again, the phone has the
worst-case specifications.
The dilemma is that the phone may not have a good site survey
tool. Certain client devices can adjust transmitter power, so as a solution you
can set this to an appropriate setting, comparable to the lowest-performing
transmitter device.
Although it is easy to compensate for transmitter power
differences, simulating receiver sensitivity is not quite as easy. To compensate
for the difference in the client receiver, you can usually adjust the AP
transmitter power (assuming the transmitter power is programmable) to
effectively reduce the signal to the receiver, emulating a lower receiver
performance.
Figure 7-2 shows several
different devices that you might use in any given site and will have to evaluate
for overall RF performances. Included in these may be bar code scanners, VoIP
phones, PCI cards, laptops with built-in radio devices, or specialty RF
devices.
To summarize, follow these steps when selecting client survey
devices:
|
Step 1. |
Analyze the client devices intended to be used in the
site. |
|
Step 2. |
Select devices that support the defined minimum data rates and
technologies (802.11a, b, g). |
|
Step 3. |
Select a device that has programmable client power
settings. |
|
Step 4. |
Take into account device antenna gains. |
|
Step 5. |
Set the power to simulate that of the lowest power device to be
used in the site. |
Note
Many 802.11b and 802.11g devices on the market only offer 15
dBm (30 mW) of power, whereas others offer 200 mW. The 802.11a product lines
vary even more widely. Be certain to do the proper research before going on site
to perform a survey.
Connectors
Most of the connectorized APs use unique connectors and vary
from vendor to vendor. As a survey engineer, it might prove helpful to build
some very short adaptor cables, allowing the use of one style coax and antenna
connector with any of the APs you might encounter for surveys. In this manner,
you can carry a single set of antennas and accessories but still connect to any
of the various APs that the customer may specify. Figure 7-3 shows connector and adaptor cable examples you
might use in site surveying.
Antennas
Maintain a
wide selection of antennas in any survey kit. Although there are certain
standard antennas, the ability to use different antennas to cover different
situations makes the difference between a professional site survey and a survey
that has been forced to work using compromises (and sometimes performance or
cost trade-offs).
Omnidirectional
antennas are, by far, the most common, and every kit should contain a
typical omnidirectional dipole antenna. This is like a universal soldier, and
can be used in many different locations. In many cases, you will need a pair of
these for supporting the diversity capabilities of the AP. Appendix B, "Antenna Radiation Patterns," shows
examples of different antennas.
A good understanding of these antennas and their particular
type of radiation patterns will help in selecting the proper antennas.
Another very common and useful antenna is the diverse
ceiling-mountable omnidirectional antenna. You can attach this antenna below the
ceiling and locate the cable routed to the AP above the ceiling. This setup
enables you to hide the AP, enables coverage similar to a dipole antenna, and
supports diversity. An added advantage of this antenna is its ease of
deployment. Installation takes about 30 seconds; you snap the antenna onto the
ceiling crossbar.
For some applications, you might have to use a slightly
higher-gain omnidirectional antenna. For this, you have several choices. When
selecting which omni to use, remember the issue of radiation pattern. When
mounting one omni in the ceiling in an inverted fashion (like a mast-mount
antenna, hanging down from the building truss), it may have a different vertical
pattern than an antenna that is designed to be mounted in that fashion. Such an
antenna is designed to be mounted to a mast, with the feed line at the bottom.
The radiation pattern has more energy below the horizontal line than above it;
and if you invert this antenna, you will have more energy radiating into the
ceiling than at the floor. An omni antenna designed for ceiling mountings is
virtually the same antenna with a different mounting design, and it does have
the same radiation pattern. When mounted to the ceiling, however, it radiates
more energy downward, as designed.
An omni antenna is almost never intended to be mounted in a
horizontal (sideways) position because the radiation pattern would radiate in a
vertical pattern rather than a horizontal pattern.
Sometimes a higher-gain omni is needed, such as the 5-dBi
antennas, but the environment will not permit (because of aesthetics) the
ceiling-mount or mast-mount antennas. A few antennas are available that offer a
"stealthy" look. The pillar-mount antenna provides a diversity antenna with a
look that blends into the surrounding walls.
Directional antennas are often used for coverage of outdoor
sites, but they can also prove very useful in covering indoor areas. One of the
most useful is a lower-gain patch or panel antenna. There are quite a few
different versions out there, but they are all similar in performance if the
gain is similar.
One advantage of a panel or patch antenna is that they can be
mounted to a wall, with the cables exiting through the wall, or up behind the
ceiling. And the antenna itself can be painted the same color as the wall. (Just
be certain the paint has no metallic properties.)
For certain applications, such as a very long corridor, a
warehouse with long narrow racks, or even certain manufacturing areas, a Yagi
can be used. It is not uncommon to use directional antennas to "fill" a
particular area with RF. An airplane hangar is one site where this has been
done. The hangar ceiling is extremely high, and the antennas must be placed
above the planes. The Yagis are mounted in the ceiling and pointed down to
provide coverage of the area below it.
No single antenna is perfect for all applications. A variety of
antennas are included in most site survey kits because they are needed at
certain times. (See Figure 7-4.) The
customer in many cases dictates antenna choice and placement. A customer might
not want the antenna to be visible or to be located in a high-traffic area. By
carrying a variety of antennas, you will be prepared for any situation. To
summarize, the recommended minimum collection of antennas should include (but is
not limited to) antennas similar to the following:
-
2.2-dBi dipole antenna rubber duck
-
2.2-dBi diversity omni ceiling-mount antenna
-
2.0-dBi ceiling-tile antenna
-
5.2-dBi mast-mount antenna
-
5.2-dBi omni ceiling mount
-
5.2-dBi pillar-mount diversity omni antenna
-
6.0-dBi diversity patch wall-mount antenna
-
8.5-dBi hemispherical patch antenna
-
10.0-dBi Yagi mast-mount antenna
-
13.5-dBi Yagi mast-mount antenna
If you plan on using diversity antennas, you need to carry two
of every antenna unless the antenna is especially made to support diversity and
contains two antennas.
Do not use a different antenna and attempt to "guesstimate" the
coverage. You are performing the site survey to take the
guesswork out of the installation.
Cables
Additional antenna cable introduces loss, and therefore
should be avoided whenever possible. In some situations, however, it is not
possible to locate the AP within reach of the antenna's attached cable. In these
cases, you can use coax cable to extend the reach.
As RF energy is carried between the antenna and the radio
equipment through a coaxial cable, it introduces signal loss for both the
transmitter and receiver. To reduce signal loss, minimize the cable length and
use only a very low-loss antenna cable to connect radio devices to antennas.
It is not necessary to carry various cables in a survey kit. If
you anticipate using a cable for the actual install, the survey should note what
cable is to be used (type and length), and the survey engineer will simulate the
cable loss by using an attenuator. This enables the engineer to simulate
virtually any cable type and length. It is recommended, however, to carry the
information for typical coax cables that would be specified. This way the
engineer can calculate the losses and program the attenuator with the proper
simulation loss.
Figure 7-5 shows the
LMR-400 and LMR-600 cables. Table 7-1
describes the features of each.
Table 7-1. Features of the LMR-400 and LMR-600
Cables
|
Cable Type |
2.4-GHz Loss (dB/100
feet) |
5.8-GHz Loss (dB/100
feet) |
|
LMR-400 |
6.6 |
10.8 |
|
LMR-600 |
4.4 |
7.25 |
It might prove helpful to carry a single 10- or 20-foot (3-
or 6-meter) cable in the kit for surveys in locations where the AP cannot be
located with the antenna, even for the survey itself. Locations such as
extremely wet areas (processing plants) or extremely cold areas (freezers) might
require the survey AP be located outside the environment and the antenna located
inside the area. In this event, you might need a cable to complete the
survey.
Attenuators
Always perform surveys using the equipment that will eventually
be installed. This can sometimes be difficult with splitters, lightning
arrestors, and extension cables.
Instead of carrying one of every length of cable, lightning
arrestors, splitters, and other accessories, some engineers outfit the site
survey kit with an attenuator. The attenuator enables you to inject varying
amounts of loss without needing the actual accessories.
The attenuator must
have the proper RF specifications. Many attenuators are only good up to 1 or 2
GHz, and you must use one that has a specification high enough to cover the band
that you will be using. An attenuator good for up to 2.7 GHz is fine for a
2.4-GHz survey, but it will not work in a 5-GHz system. An attenuator with a
6-GHz rating could be used at either 2.4 or 5 GHz. Figure 7-6 shows a variable attenuator that can be used to
simulate various losses in the antenna system or cables.
Attenuators are expensive. As frequency of the device goes
up, so does the price. This is one area where you cannot cut costs.
Physical Measuring Devices
One tool that is often overlooked is a measuring
devicesomething that can measure long-distance runs of hundreds or even
thousands of feet or meters. In many cases, you need to add measurements to your
drawings for room size, building size, or even parking lot size. Or you might
need to measure the distance a cable will have to run between and AP and a
switch (keeping in mind whether that will meet the intended cable-length
limitation). If you guess the Category (Cat) 5
cable run to be 300 feet (91 meters) and it turns out to be 380 feet (116
meters), the customer will be very dissatisfied. For some measurements, a tape
measure might prove impractical and difficult to use (especially for one
person). Your kit also should include a measuring wheel similar to the one shown
in Figure 7-7.
Many survey engineers find that some of the more advanced
equipment, such as laser measuring devices and range finders, can save time and
effort. For measuring vertical distances, these can prove extremely helpful,
especially when doing a survey alone. An alternative method is to use a rope
marked in 10-foot (3-meter) increments so that you can accurately judge
distances floor to ceiling.
RF Analyzers
In today's world of communications, RF is everywhere, in every
possible environment. Therefore, survey engineers must look at the RF
environment for other radio devices and interference. Failure to do so can
result in a system that is installed but never works.
Looking at the RF spectrum requires a specialized receiver or a
spectrum analyzer. The next section examines different tools for analyzing the
RF environment.
Spectrum Analyzer
The spectrum analyzer is the best and most useful tool for
RF spectrum analyses. It not only provides a look at the spectrum in which the
WLAN system will operate, it can also determine the frequency of other signals,
the duration they are transmitting, the strength of the signal, and with some
experience the engineer may even be able to determine the type of modulation in
use, leading to an understanding of the type of device being used. By using a
directional antenna, you can use a spectrum analyzer to locate a signal
source.
Spectrum analyzers have several drawbacks as a site survey
tool. The first drawback is cost. For an analyzer that supports the 5-GHz band,
the cost could reach well above $10,000! Second, most spectrum analyzers require
AC power. A few models are offered with battery packs, but the duration of
operation is limited based on the battery power.
Portable Analyzer Tools
Portable analyzer tools such as the Berkley Valtronics
Grasshopper can prove very useful and cost-effective (see Figure 7-8). The Grasshopper is a handheld, wireless
receiver designed specifically for sweeping and optimizing WLANs. It can measure
coverage of 802.11b systems, enabling the user to measure and determine the AP
and received signal strength identification
(RSSI) signal levels (which helps the user locate the APs throughout a
building). It can also detect and differentiate from narrowband multipath
interferences (such as from microwave ovens and frequency-hopping systems) and
features a built-in display, keypad, and removable battery pack for true
portability.
Some WLAN products enable you to sniff, or listen, to the
RF, acting like a spectrum sniffer listening to frequencies the gear is designed
to operate in. This is a convenient tool, but it is not as effective as a good
spectrum analyzer. One such WLAN product is the Cisco Aironet AP1200 AP. It
offers a feature called carrier test. It provides
a dBm level of the highest signal received on any of the available 802.11
channels. The AirMagnet wireless sniffer product also offers a feature to sniff
the air for other RF energy. Although these tools are very useful, they do not
compare to the versatility of a spectrum analyzer in the hands of an experienced
engineer.
Two-Way Radios
In some situations, two engineers will conduct a survey
together. In such cases, a communication link between them can help to reduce
time. Family Radio Service walkie-talkies are very inexpensive and typically
require no license (at least in the U.S.). Before using them in a customer site,
however, always inquire as to any company regulations about radios. Some sites
may have areas where high-power two-way radios are not permitted.
Outdoor Tools
If the survey is for outdoor links, you might need other
equipment. Tools to determine necessary antenna height, waterproof-test
enclosures (for the equipment during testing), and Global Positioning System (GPS) receivers are just a
few tools that can prove very helpful. Chapter 14, "Outdoor Bridge Deployments," covers some
of the tools used in outdoor site surveys.
Battery Packs
APs require power to operate. However, power will not always be
available nearby while you conduct a site survey. A solution that enables you to
power the AP without running long cables is the best method for surveying. A
good survey battery pack will last for at least eight hours, enabling the
engineer to survey all day without having to recharge.
Because power requirements vary from one vendor AP to another,
a single power pack is not something that is universal. Some battery packs are
available with various power offerings (such as the TerraWave pack shown in Figure 7-9).
With this particular site survey battery pack, you can easily
survey for up to eight hours without worry about having to pull around a long AC
power cable or worry about having anyone trip over your power source. It
provides 5V DC and 12V DC outputs for powering Telxon, Aironet, and some Cisco,
Lucent, Symbol, Intel, and Proxim APs and bridges. With the in-line power adaptor, you can
use that supply to power other APs that utilize the Cisco Power over Ethernet (PoE) scheme or 802.3af PoE.
Another way to provide power to APs is to use a UPS type of
supply that provides AC power. However, these can become very heavy, and their
overall available power is usually fewer than eight hours.
Also recommended is a fast charger for the client site survey
tool. If a laptop is used, spare battery packs that can be charged separate from
the laptop are always recommended. Wireless PC cards require a constant source
of power while surveying and might reduce battery life to fewer than two
hours.
Another possibility is a high-capacity battery pack such as the
N-Charge power system from Valence Technologies (see Figure 7-10). This type of battery can provide power to a
typical laptop for four to five hours. (The specification says up to 10, but
that does not factor in running a radio in site survey mode.)
Digital Camera
Another device to consider adding to your site survey kit is
a digital camera. Taking a picture of antenna mountings, AP locations, special
fixtures, or locations and adding these to the site survey report can save
countless hours trying to describe details in words.
Use pictures whenever you have any doubt about how to mount or
about where to locate something. Also use pictures to show possible interference
(environmental) issues in the survey reports.
Mounting Hardware
Mounting hardware for the site survey can be quite a bit
different from what is needed for the actual install. For an install, you are
leaving the WLAN components installed permanently. For the site survey, the
mounting is just long enough to complete your measurements. However, based on
the environment and who is around when you are surveying, mounting options can
vary widely. Although the installation might be temporary, it needs to be secure
enough so that something does not come loose and fall, creating a hazard for
someone who happens to be in the area while you are working. It must also be
secure enough that someone does not "walk off" with the test gear while you are
making your rounds performing testing.
Always carry an
AP mounting bracket (when available). Your kit should also contain various
mounting solutions for the bracket (beam clamps and C-clamps, for example) as
well as mounting brackets for each antenna (when available). Ladders and a
collapsible pole for temporary placement of an AP prove extremely useful as
well. Beyond this, you must again be creative. Zip ties, duct tape, bailing
wire, electrical tape, two-sided tape, and Velcro are common components in a
good engineer's kit (see Figure
7-11).
Remember, during a survey there is no bad mounting solution
except the solution that does not properly secure the AP, battery pack, and
antenna. If these are not properly secured, not only might you damage your
equipment, you risk injuring yourself or others when the equipment comes
crashing down.
Sections
