Additional Base Stations

Once network operators have secured central office facilities, they must then turn their attention
to the matter of access points. Initially, the network may have but a single access point that
will be situated within or adjacent to the central office, but, as the network expands, additional
access points will be required. Determining precise requirements in this regard is no simple
matter, but it is essential to the success of the network.
Several factors should play a role in the optimal siting of access points in the network,
including the nature of the equipment (line-of-sight or NLOS and beam-forming antenna, or
lack thereof), the effective range of the access point transmitter, the number of potential customers
within range of the transmitter, the extent to which frequency reuse can be achieved
within a given area, and the availability of suitable locations at an affordable price. None of
these factors should be considered in isolation.
A network based on a large number of NLOS subscribers is going to have to put the subscribers
closer to the base stations—in other words, more base stations will be needed. Density
of the subscriber population and the reach of the transceivers will also have a bearing on where
base stations are situated. The first Multichannel Multipoint Distribution Service (MMDS) networks
that transmitted over licensed spectrum at relatively high power and could maintain
links over distances exceeding ten miles could serve thinly scattered customers over a large
area with a single base station. Network operators using unlicensed spectrum and transmitting
at 1 watt maximum power are unlikely to be able to duplicate that and instead will require a
number of base stations for the same area and the same customer density. I want to generalize
in this regard and say that at such-and-such customer density over so many square miles at 5.8
gigahertz (GHz) this many base stations will be required, but it is never that simple, because
the same equipment may perform differently in different radio frequency (RF) environments.
Wireless network operators add base stations reluctantly—doing so only when the reach
and/or capacity of the existing base stations is insufficient to serve what is seen as a sizable
number of potential customers. One knows when one is approaching that point because of
increasing network congestion, and one then sets out to find a location of the new base station.
Insofar as possible, the network operator should attempt to plot traffic patterns at least two
years into the future, however, and should plan the location of future base stations long before
they are actually required.
This brings you to the final topic regarding base stations: identifying suitable locations for
base stations.
Particularly in the case of line-of-sight equipment, one wants a base station with a minimum
of obstructed pathways to potential customers. In most cases, the base station antenna
should occupy an elevated position where sight lines extend well over the tops of the tallest
intervening structures.
In recent years the tendency in American cities has been to place antennas for wireless
networks on towers specially constructed for the purpose and owned by companies such as
American Tower and SBA, which have developed businesses based on the lease of space to network
operators. The rise of such companies has been the direct result of the proliferation of
mobile telephone cell sites and the determination on the part of municipal governments to limit their number. By obliging commercial operators to collocate on few approved sites, that
proliferation has been checked, and, coincidentally, a new industry has emerged.
Broadband wireless operators can certainly elect to occupy such a site, but they may not
necessarily be welcome or able to afford the lease if they are. Tower owners do not charge uniform
rates, but a rough average is $1,000 a month. This is nothing to a cellular operator with
tens of thousands of customers, but for a struggling wireless broadband operator it is a significant
expenditure especially when that operator is expected to sign a 30-year lease in the
bargain. Another issue with towers is that they are primarily intended for hosting lowerfrequency
mobile services rather than networks operating above 2GHz, and they may not
provide optimal elevation since the mobile services lack line-of-sight requirements.
If the broadband wireless network operators are disinclined or unable to gain access to a
tower, they can always attempt to negotiate roof rights on a multistory building. Here it is difficult
to generalize, because rates can vary tremendously. In some instances, if the network
operator is willing to provide data services to the tenants of the structure in question, recurrent
fees may be reduced or even waived.