Performing Site Surveys and Determining Link Budgets

A site survey is simply a determination of the extent of the interference levels in the frequencies
of interest and of the severity of obstructions and multipath at the precise locations in which
one wants to install terminals. Such a survey enables the network operator to determine if
unacceptable amounts of interference are present at various locales such that adequate signal
quality cannot be obtained, as well as the extent to which blockages restrict blanket coverage
in a given area. Ideally, the site survey should be an ongoing process and should encompass
every link in the network. In practice, something considerably less extensive will be performed
because of the prohibitive cost of deploying technicians at every location.
Site surveys are performed manually and painstakingly by taking innumerable measurements
with a radio spectrum analyzer and eventually with an actual radio transceiver. It is
grunt work, but a certain minimum surveying is required for informed network planning. It is
also one of the first steps a network operator should take, and if unlicensed spectrum is to be
utilized, it should be perhaps the second step after a determination has been made as to what
spectrum is available.
In theory, a site survey should encompass every location where either an access point or a
subscriber terminal may be situated, but of course the time and expense required to achieve
that theoretical ideal are enormous in the case of a large metropolitan network. Therefore, almost inevitably network operators will have to satisfy themselves with a less-than-thorough
accounting of the wireless environment in which they will be operating. At the least, site surveys
should be done for all base stations and for all subscriber terminals serving multiple
customers such as terminals at business parks and high-rises, and multi-tenant residential
buildings.
In the case of NLOS radios and mesh architectures, the importance of the site survey may
be lessened. After all, the intent of both approaches is to foster self-installation, and that virtually
precludes a site survey except perhaps through some simplified software utility included
with the subscriber terminal. I have already touched upon meshes and have indicated the
many uncertainties associated with their deployment and the plethora of paths over which an
individual transceiver may transmit. Obviously a comprehensive site survey within a mesh
network would be nearly overwhelming. Chapter 5 covers NLOS approaches extensively,
including issues of terminal siting, but here you should observe that NLOS self-installed equipment
is generally placed indoors, which greatly reduces the maximum permissible distance
from the base station and, if generally used throughout the network, may require an increase
in the number of base stations over that of a deployment characterized by professionally
installed outdoor antennas. Nevertheless, some will argue that line of sight always entails more
base stations in the end in order to get around blockage.
If an extensive number of indoor installations is under consideration, the network operator
may be forced to weigh the expense of more base stations relative to more truck rolls. I have
already mentioned the cost of truck rolls. A truck roll involving not only installation work but a
complete site survey may take the better part of a day and may cost the network operator hundreds
of dollars. If such truck rolls are required for most customers within a large metropolitan
network, and the customer population eventually grows into the thousands, they will represent
a huge operational expense, one made worse if the network experiences a lot of churn. On
the other hand, base stations can entail considerable up-front capital expenditures and significant
recurrent costs. Within the lower microwave regions a base station may carry an up-front
cost as high as $10,000, a backhaul cost of $1,000 per month, and a site lease cost of another
$1,000 per month (these are typical figures, but by no means benchmarks) for a total of $34,000
during the first year of operation, not counting utilities and allocation of staff time. On the
other hand, 100 truck rolls performed within a single cell could exceed the total expenditure on
one base station.
I prefer NLOS solutions because they enable the network operator to reach customers who
would otherwise be inaccessible as well as reduce the number of site surveys that must be completed.
One should understand that NLOS is not a panacea, however.
To return to the discussion of actual site survey procedures, note that the person conducting
the site survey is seeking three kinds of information. First, he is looking at the amount of
interference in a given band and adjacent bands either from existing users or from accidental
interferers. Second, he is attempting to determine what the signal strength will be at the
antenna at the site in question for signals emanating from various other points in the network.
Finally, he is looking at the integrity of the data stream and such factors as bit error rate, jitter,
latency, and throughput rate.
Although currently available instrumentation is precise, there is still apt to be a
certain amount of guesswork in doing the survey because one can almost never determine
the concentration of customers in given areas covered by a base station, and course
interference levels can change over time, sometimes markedly. The initial site survey does provide the network operator with baseline information, however, and as such it is an indispensable
part of the planning process.
A network operator can elect to use internal staff to perform the site survey or use outside
consultants. In either case prior experience is highly desirable. At a minimum the individuals
performing the survey must be familiar with the basic concepts discussed in the next section.
While the survey is under way, all observations and readings should be carefully and completely
recorded, and the final report should be kept in a secure place where it can be consulted
by responsible individuals in the organization as the need arises. The report should indicate
the exact locations of all network nodes, including elevations. Using the Global Positioning
System (GPS) coordinates to record positions is an excellent idea.