I'm trying to help my aunt set up a wireless network on her ranch in a
remote corner of New Mexico, and have run into a number of questions
about outdoor and long-range WiFi that I hope people can help me
answer.
The central office building has Internet access via DSL, to which we
would like to provide access from other locations around the ranch.
Digging two thousand feet of trenches and laying shielded fiber would
be prohibitively labor intensive and expensive, so wireless seems like
the logical answer.
Geography
Many of the buildings and outdoor spaces where we'd like to have access
are clustered within 150 feet of the main office. However, a few key
buildings are spaced further away, up to approximately 600 feet, and
one has quite a few trees between it and the main building.
In the map below, the main office is marked with a red border, while
the primary other buildings we'd like to provide access to are shown
with a blue border. The blue-marked building in the upper-right is
completely blocked off from the main office by an area of many trees,
but has near line-of-sight to the workshop, which is marked with a
green border.
http://matthew.cavalletto.org/wifi_outdoors.html
Other Constraints
There will only be a few simultaneous users, so I don't expect that
bandwidth contention will be an issue.
There are a fair number of trees on the property, which will attenuate
the signal.
On the upside, the nearest neighbors are quite distant, so there
shouldn't be any interference with other wireless devices.
All of the computers being used are Macs of various vintages, and
there's no on-site tech guru, so I'd prefer to use Apple's AirPort base
stations for simplicity, but we could live with third-party access
points if needed.
We don't have a fixed budget, but if it looks like the price is going
to be much over $1,000 they're likely to just settle for getting
another DSL line or satellite-based Internet access at one or two
additional locations.
Ideas and Questions
The first step would seem to be placing a WiFi access point in the main
office, connected to the DSL router, with an outdoor omni-directional
antenna on the roof to provide access to the main cluster of buildings.
- Which model of outdoor omni-directional antenna should we use on the
roof of the main office?
- Does it make sense to get an amplifier to increase the signal
strength?
For the building shown on the left side of the map, which is about 250
feet from the main office and has some intervening trees, I think we
may need a directional antenna pointed towards the main building.
- Instead of getting an outdoor antenna, can we just use an indoor
directional antenna and point it through a window, or does that
significantly attenuate the signal?
The building in the bottom-right corner of the map, is even further
away: about 6-700 feet.
- Is a directional antenna likely to be sufficient over this kind of
distance?
For the building on the upper-right corner of the map, I'm concerned
that even a directional antenna may not be able to punch through all of
trees that lie between it and the main building, so I've been thinking
of placing another access point at the workshop, which has
line-of-sight to both the main office and the remote building, and
using that to propagate the signal.
- Is a directional antenna on the remote building likely to be
sufficient, or do we need a directional antenna on each end of the
link?
- Is there a decent wireless-to-wireless bridge with two antenna ports,
so we can connect an omni-directional antenna to one to pick up the
local signal, and a directional antenna to the other to connect to the
far side of the property?
Suggestions Welcome!
I've spent a while Googling, which reinforced my initial impression
that it's difficult to predict the real-world performance of wireless
networks.
It may be that our only choice is to engage in a bit of
trial-and-error, starting by placing the initial base station and omni
antenna, then wondering around the property to see what level of signal
strength we get, and figuring out how to expand from there.
However, any suggestions or feedback people might have would be much
appreciated. Have you dealt with these kinds of issues in the past? Any
recommendations for specific equipment, general approaches, or avenues
of further research?
Thanks!
-- Simon