Determining Line of Sight and Computing Fresnel Zones

 
The site survey begins with visual observations, and the first task is to attempt to establish clear
line of sight from the subscriber location to the base station; the assumption here is that right
of way has already been granted for antenna placement in both locations. The process is quite
straightforward and involves nothing more than standing at the level of the transceiver
antenna in one of the locations and training a pair of binoculars on the other antenna site. If
a clear path lies between the two points with several yards separating the nearest obstruction
from the imaginary line connecting the two points, then one can assume that line-of-sight preconditions
have been met.
Another method available in the United States, though not always elsewhere, is to purchase
three-dimensional aerial maps of the area one wants to serve. These provide elevations
of all buildings and natural obstructions such as hills and bluffs. Arizona-based AirPhotoUSA is
a vendor that has served many wireless operators.
If obstructions directly cross the imaginary line, then one obviously lacks clear line of
sight. If obstructions nearly cross that line without quite impinging upon it, then one must proceed
to the next phase, the calculation of what is known as the Fresnel zone.
By strict definition, Fresnel zones consist of an infinite series of concentric rings surrounding
the nodal point of transmission, with each ring defined by the phase relationship between
the main beam of the transmitter and the two dominant side lobes (these terms are explained
shortly). The phase fluctuates from one zone to the next so that side lobe reflections are sometimes
in phase with the main beam and sometimes antiphase to it.
So how do such side lobes arise? Directional antennas of the sort generally employed in
broadband wireless networks tend to focus energy into narrow beams, but they never produce
just one beam. Rather, they produce a dominant central beam and two or more side lobes that
contain less energy than the main beam. They also produce low-energy rear beams. Depending
on the polarization of the transmitter, the so-called side lobes may be above or below the
main beam or on either side of it.
Now, even when the direct signal has a clear path to the receiver antenna, these side lobes
may encounter obstructions since they are offset from the main beam and describe different
propagation paths. When they do strike obstructions, they will be at least partially reflected,
and the reflections may impinge upon the main lobe, either reinforcing or canceling the signal
depending on the phase relationship of the two.
Here an element of confusion enters the discussion because field technicians tend to
employ a different definition of Fresnel zones. By this second definition, a Fresnel zone refers
not to a concentric ring but to an elliptical volume surrounding the imaginary line of sight and representing a region where the presence of physical obstructions will set up strong side lobe
reflections that may disturb the main beam.
The reflected signal may, as you have seen, be in phase or antiphase to the main beam, and
thus it may reinforce it or partially cancel it. One may think that reinforcements would be preferable
to cancellations, but in fact both are undesirable, especially so within the first Fresnel
zone (following the earlier definition) where strong early reflections of the side lobes occur.
The network operator simply has to plot a path that is free from major obstructions.
The distances involved in Fresnel zones are frequency dependent and are also a function
of the radiation patterns of the antenna used. It is best to consult with the manufacturer of the
subscriber terminals and base station equipment to determine the extent of the area above and
below line of sight that must be kept clear of obstructions. Most commercial programs for cell
site location include subprograms for Fresnel zone calculation, and Proxim, a leading manufacturer
of broadband wireless equipment, makes a well-regarded stand-alone program, so
one need not despair if one’s math skills are wanting.
If obstructions do stand between the base station and a valuable subscriber site such as an
office building or residence with multiple subscribers, one remedy may be simply to raise the
base station antenna on a mast at an elevation where it is well above any obstructions. Such a
tactic must not be regarded as a perfect solution, however, because an antenna that is too high
will not be able to reach subscribers in the immediate area of the antenna.
It should be further noted that not all obstructions are equivalent and that considerable
differences may exist among obstructions of the same general type.
A single tree for instance may impose around 15dB to 20dB of signal loss depending on
type and size. A grove of trees may up that figure to 30dB. A building may represent a total loss
of 30dB while a low hill could exceed 40dB. A truly interesting situation occurs when trees sway
in the wind. Momentary variations in loss may exceed 10dB, and generally one must design
around the worst case.
Bear in mind that the presence of obstructions within the first Fresnel zone does not preclude
the establishment of an airlink. It just means that more transmit power will be needed to
achieve the same signal integrity as in an unobstructed path.