Installation guide

79
Wi-Fi Location-Based Services—Design and Deployment Considerations
OL-11612-01
Deployment Best Practices
Location Tracking Challenges
Outdoor Environments
Outdoor wireless deployments tend to be much different from the types of indoor deployments that have
been described thus far. The deployment best practices described in this document do not readily lend
themselves to easy deployment outdoors. The access point densities, inter-access point spacing, and
antenna heights discussed, although acceptable for indoor deployments, tend to make outdoor location
deployment less than optimal.
Because of this as well as other factors, the use of the location appliance for location tracking of mobile
devices or RFID tags in an outdoor environment is not recommended at this time. The performance
specifications noted in Accuracy and Precision of the Cisco LBS Solution, page 25, apply only to usage
of the product within indoor environments.
Non-Uniform Environments
When an RF calibration is performed, the information gathered is used to make certain assumptions
about the overall attenuation present in the environment. These assumptions are applied to internal
distance calculations using a path loss model that is applied to the entire floor. With proper design and
deployment in environments with uniform construction throughout, this approach performs rather well.
However, in some cases the network designer is faced with challenges because of an environment that is
of non-uniform construction. An example is a single floor consisting of cubicles in one area, metal
racking and electronic equipment in a second area, and a large group of individual offices with concrete
block walls in a third. The application of an overall path loss model to a floor with clearly non-uniform
attenuation characteristics poses challenges to obtaining optimum location accuracy and precision. This
is primarily because each of the three areas exhibit different degrees of attenuation and other propagation
anomalies.
The network designer facing this type of environmental challenge may choose to do one of the following:
• Option One—If location performance is of equal concern in all concerned areas, perform a
calibration on the entire floor as usual, with the understanding that the resulting overall path loss
model may not be optimally tuned to the particular characteristics of any individual floor area. This
approach “averages out” the differences between the various areas on the floor and computes a path
loss model for the overall floor.
• Option Two—If location performance is of more concern in one of the areas as opposed to the
others, perform the calibration focusing only on the area of primary concern but applying the
resulting calibration model to the floor as a whole. An example of this type of deployment might be
a location that combines a very large stockroom with an office environment, where the emphasis on
location performance is in the stockroom.
• Option Three—Define each of the areas of the floor as individual “sub-floors” and define each of
these sub-floors to WCS and the location appliance as if they were physically separate floors. A
separate calibration is then performed on each of these floors and a customized calibration model is
applied to the floor. The end result is that each sub-floor area is treated as a separate floor for the
purpose of location tracking.
Each of these choices has its advantages and disadvantages. The first and second options offer more
easily manageable and recognizable alternatives, especially when dealing with multi-story facilities.
Under the first option, the path loss model computed is less than optimal as it relates to any particular
area of the floor in general; however, the level of variation should be less than in the second option. The
second option, although potentially offering more variation in performance between different areas of
the floor, does present the capability for superior performance in areas of the floor that are of the most