|
Automatic and Assisted Site Survey Tools
Jan 09,2007 00:00
by
admin
Automatic and Assisted Site Survey Tools
Automated surveys include both assisted and totally
automatic surveys. As suggested earlier, the assisted survey is really a mix
between a manual and a totally automatic survey, providing benefits of both.
Here the discussion focuses on only one of each type, using the Aruba Networks
to discuss automatic surveys and the Cisco SWAN framework for discussing the
assisted survey. Although the actual implementation of these types of surveys
will vary somewhat between vendors and is unique to each product type, they are
similar. The intention of this section is to give an idea of how they work in
general terms, and the pros and cons of each type.
Aruba
Aruba Networks offers a WLAN system that includes a network
switch acting as the WLAN controller, and APs that are managed by the switch.
When installing an Aruba network, you must first install the APs and WLAN. The
initial reaction to an installation like this is often, "Where do you place the
APs? Isn't that what a site survey is for?" The answer is to use a best-guess
process based on user density and location (similar to the coffee cup survey
discussed earlier), and to place enough APs to feel certain there will be more
than enough RF coverage in the desired areas. This means using more APs than
might be necessary based on user density. This overengineering enables you to
cover any RF holes that might result from RF shadows and multipath
interference.
After
installing the network and APs, perform a calibration of the APs. The calibration is a process typically performed first at
the time of network installation and then periodically as the physical
environment changes significantly. Normally this is performed on a per-building,
per-band type (2.4/5-GHz) basis. All 802.11a radios in a given building are
calibrated at the same time, and all 802.11g radios are calibrated at the same
time with negligible impact on the availability of the WLAN.
During calibration, all APs and access
monitors (AMs; AP type of device used only for monitoring and not for
communication to clients) communicate with each other at the various data rates
and different transmit power levels. This process allows the Aruba switch to
build an RF-based map of the network topology, learning about environmental
characteristics such as attenuation, interference, and reflection. When
calibration is finished, Aruba's Wi-Fi switch automatically configures AP/AM
mode of the APs, transmit power levels, and channel selection to minimize
interference and maximize coverage and throughput.
Because such a system is not truly bullet proof and there is a
high probability of RF holes, Aruba provides a feature called Coverage Hole
Detection. Coverage Hole Detection looks for clients that have been unable to
associate to any AP (not sure how this feature works, because the device is not
associated to an AP and therefore the AP does not know the client is even
there), or associating at low data rates or low signal strength. These symptoms
indicate areas of a building where holes in RF coverage exist. When the system
detects such coverage holes, the administrator is notified of the condition via
the event log. At that point, the administrator must go into the facility and
place another AP.
Cisco Assisted Survey Utility
The Cisco Assisted Site Survey tool is a cross between the
automated survey and the manual survey. In most cases, there is a need to
perform some initial testing at the site before using the WLAN manager site
survey tool, and hence the name Assisted Site
Survey. The tool is a part of the Wireless LAN
Solution Engine (WLSE).
The first step is to determine the area (using the floor plan)
that will cover the desired number of users. This must be done in the different
types of areas where densities will vary, such as conference rooms, lunch rooms,
and so on, because these require a higher number of APs to provide the ratio of
users to APs specified in the design.
After determining that the cell size of a full-power AP, at the
defined minimum data rate, is greater than needed to provide coverage for the
maximum number of users for a single AP, you can use the floor plan to determine
locations for the APs. Some areas (such as conference rooms) will use a smaller
cell size to maintain the proper ratio of users to APs. After the AP locations
have been determined, the installation must be done. After installing the APs,
you can begin the assisted site survey.
If there are
areas for which the maximum power and minimum data rates cell size are
insufficient to provide coverage for the proposed ratio of users to APs, a
manual site survey will provide a better economics of scale for the network.
Manual surveys enable you to more accurately determine cell sizing, select the
appropriate antennas, and accurately provide coverage while using the minimum
number of APs. If some areas use unusual antennas (such as the Yagi), a manual
site survey provides more guarantees for coverage in desired areas.
The Cisco Assisted Site Survey tool requires the use of the
Cisco WLSE on the network, and the survey is completed in three parts. The first
part is an AP scan, which takes control of all APs and performs a scan by each
AP, listening for all other APs. When this is completed, there is a mode (it is
optional, but highly recommended) known as the client
walkabout. This is where a Cisco client (or Cisco CE client) is walked
throughout the facility, usually around the perimeter, and throughout the
internal area. The client in turn reports back to WLSE regarding the signal
levels it reads from the APs as it moves about.
In the final step, radio-management-assisted configuration, the
WLSE uses the data information from the client walkabout and AP scan to adjust
to the signal levels reported by the clients at each data point. WLSE then
increases or decreases the transmit power of the AP to provide an adequate S/N
ratio.
Clients in most applications are 3 feet to 4 feet off the floor
and are generally subject to a good deal more attenuation than an AP. For this
reason, the WLSE optional client walkabout mode will factor in the results of
the survey data collected by the clients, providing a much more robust survey
than one that uses the AP scan only. Another factor that is missed in an
AP-scan-only type of utility is that the S/N ratio changes depending on the
modulation type and data rate (and depending on the location in the facility and
even within a single cell). The client walkabout takes this into account.
|