WAP11 Bridge
Disclaimer: The following page documents my experience modifying hardware to do a job that it
was not originally intended to do. If you decide to attempt a similar project to the one
described on this page, you do so AT YOUR OWN RISK. I take no responsibility for any of the
information given on this page. Be warned that some of the risks involved in this project may
end up destroying your network gear, cause you to get struck
by lightening, burn down your house, or microwave your neighbor's cat.
This page describes my efforts to connect two homes in Santa Clara, CA using Linksys WAP11 Access Points
in bridge mode. The houses are relatively close together - about 400 meters. As far as bridging goes,
this is a small distance from what I've read, but it's longer than what the WAP11 is technically capable
with it's stock antennas. The reason for the bridge is because one house is out of range for DSL and
cable modems, but the nearby house has cable modem service through AT&T broadband and offered to
share the service free of charge.
Parts
- Linksys WAP11 v2.2 (no longer sold... try WAP54)
802.11b Access Point - $130 x 2 = $260
- Linksys BEFSR41
Cable/DSL Router w/ 4-Port Switch - $80
- Generic 5 Port 10/100 Switch - $30
- Hyperlink Technologies HG2412Y-RTF
12dBi Enclosed Yagi Antenna - $50 x 2 = $100
- Pelican APP-1200E Waterproof Case - $26 x 2 = $52
Theory
Much of what I've learned about data communication in the 2.4 GHz ISM came from reading
this
VHF/UHF/Microwave Radio Propagation Primer for Digital Experimenters.
I highly recommend it to anyone interested in extending the range of 802.11b equipment. Also, much of
the math involved with radio communication involves calculations in dB. dB by itself describes a relationship
between two values and has no value or meaning by itself. Every 10dB represents a magnitude of 10. Also, every
3dB represents just slightly less than a magnitude of 2. In radio, the units dBm and dBW are often used
to describe signal strength relative to 1mW and 1W respectively. For example, 0dBm = 1mW, 3dBm = 2mW,
6dBm = 4mW, 10dBm = 10mW, 20dBm = 100mW, and 30dBm = 1000mW = 0dBW = 1W. All of the signal strength calculations
for my project were done in dBm.
The Linksys WAP11 v2.2 access point claims an output power of 20dBm (100mW) in it's
User's Guide (see Appendix D: Specifications - page 44).
According to the FCC
Grant for Equipment Authorization for O7J-GL2422AP (the FCC ID of the wap11 v2.2), the output
power to the antenna is 0.038W = 38mW = 15.8dBm. (If you're still curious, the list of
published FCC documents is here.) The FCC documents also state the maximum
gain of the dipole antennas shipped with the wap11 v2.2 is 5dBi, so this is probably why the user's guide
claims a power output of 20dBm. Since we won't be using the stock antennas, we'll use the 15.8dBm figure
as the power output at the connector of the wap11.
The wap11's documentation doesn't specify the receiver sensitivity however I was able to locate another
device which uses the RFMD radio chipset
and claims to deliver 11, 5.5, 2, and 1mbps at -82dBm, -85dBm, -90dBm, and -93dBm respectively.
The free space path loss through the air can be found by Lp = 32.4 + 20log(f) + 20log(d) dB, with f being
the frequency in MHz, and d being the path distance in km. In my case, f is 2437 MHz, and d is 0.400 km,
giving a free space path loss of Lp = 92.2dB.
In addition to the free space path loss, there is also some loss associated with the connectors and
12" of cable on each antenna. The 12" of cable on the Hyperlink Yagi antennas is LMR195, which has a
19.0dB per 100' attenuation at 2500MHz. This means for each antenna there is an attenuation of 0.2dB
for the cable. The antenna connects to the wap11 with an RP-TNC connector, which at 2.4GHz has a
maximum attenuation of 0.2dB. Adding
the connectors and cables together we get an total additional loss of 0.8dB.
We can figure out the power at the receiver using the equation Pr = Pt - Lp + Gt + Gr - Lt - Lr, where Pt
is the transmission power, Lp is the path loss, Gt and Gr are the antenna gains, and Lt and Lr are the additional
path losses at the transmitter and receiver. Using the values from above, we get Pr = 15.8dBm - 92.2dB + 12dBi + 12dBi
- 0.4dB - 0.4dB = -53.2dBm. According to the receiver sensitivity, we can reach the maximum data rate at -82dBm and
above, giving our setup a safety magin of 28.8dB. According to these figures there should be no problem at all
achieving a full 11mbps link.
We can also calculate the theoretical maximum path distance by solving for Lp in the equation above.
For an 11mbps link we need Pr >= -82dBm so: Lp = Pt - Pr + Gt + Gr - Lt - Lr =
15.8dBm - (-82dBm) + 12dBi + 12dBi - 0.4dB - 0.4dB = 121dB. Turning around the free space path loss
equation to solve for distance gives us d = invlog((Lp - 32.4 - 20log(f))/20) = 11km. So, in theory
we should be able to support an 11mbps link over any distance <= 11km (7 miles). A 1mbps link requires Pr >=
-93dBm which requires Lp <= 132dB giving us a theoretical maximum distance of 39km (24 miles)... And
this is with using 12dBi Yagi antennas. Parabolic dish antennas can achieve gains in excess of 24dBi - you
do the math.
Network Design
Here's a quick overview of the network design plan:
Enclosures
The wap11 is placed in a sealed enclosure mounted to the antenna mast on each side. This
keeps the lengths of coax antenna cable to a minimum. The stock Yagi antennas come with a 12" pigtail on them.
Power is fed to the enclosure over the two unused wire pairs in the cat5 cable. The stock wap11 power supply will
be located indoors near the hub or router the wap11 connects to, and the 5V power will be spliced into the cable
going outdoors to the wap11.
The enclosure I chose for this project is the
Pelican APP-1200E. I got the idea to use
this enclosure when I saw this web page
(link now broken)
describing a solar powered wap11 setup (although it appears like they're using the wap11 version 1). The
APP-1200E is a water/air tight case. In order to mount the wap11 hardware inside the case I had to drill
a few holes in the case which had to be sealed with Silicone Caulk. I decided to remove the hardware from the
wap11 and mount it directly into the enclosure. I did this by using small 1/8" diameter nuts to go through the
back of the case, through a 1/4" plastic spacer ring, and then through the mounting slots on the internal
metal frame of the wap11. I mounted the board such that the left antenna feed could be removed from the
wap11's metal frame and mounted in a hole in the top of the enclosure. I also drilled holes in the back of
the enclosure which will be used to mount the enclosure on the antenna mast with u-bolts. Again, all of these
holes in the enclosure have to be sealed once the enclosure is mounted on the antenna mast to avoid moisture
from ruining the electronics.
Results
A few days after receiving all of the various components the bridge was up and running as expected.
It turns out that the path between the houses was not exactly line-of-sight, as there were a few
trees in the way, but that didn't seem to hamper the performance too much. The link holds steady
at 11mbps with a packet loss of around 1 1500b packet out of 500. Not too bad.
Pictures
Coming soon...
| 
