The secret is to cut the losses as much as possible. The best way is to mount the access point or bridge right at the antenna in a weatherproof box and feed it with outdoor cat 5 cable using power over Ethernet (POE). The RF jumpers are short and the loss is minimized. It may be of some benefit to use horizontal polarization on your path, most WIFI and ISM signals tend to be vertically polarized and the cross pol may provide you with a benefit. You may find that you have plenty of signal but not much throughput, this can be caused by the acknowledgement (ACK) window being too short, what happens is the propagation delay on a long path causes the receiver to stop listening before the packets arrive. The long range radios have an adjustable ACK time to prevent the problem. I like the Linksys WET-11's, they are a bridge that can be configured with an access point (AP) or a pair of them in a peer to peer mode. They are inexpensive and readily available on the used (e-Bay) market. Another interesting item is the Ubiquiti LightStation modules, they come in both 2.4 and 5 GHz models. http://www.ubnt.com/litestation2.php4 . The 5.8 GHz band is actually better for point to point use, it is less crowded and the antennas are more directional. I have a 5.8 Path that is 42 miles long and it is extremely reliable but I am using some 10 and 8 foot dishes to get the gain up and I have yet to catch it fading. I recommend that you stay with 802.11b or a, forget about the g as the receiver sensitivity goes down with extra bandwidth and they become more prone to fading. 11 MB/s is fine for most of what you need and it is faster than DSL or most cable modems.
Jack K6YC No I haven't tried it yet. I did some modeling and with 150mw, 20dbi antennas and the antenna at the friends house mounted up 21 feet (he doesn't have a tower and isn't a Ham) and the antenna mounted at least 60 feet up my tower it showed a good path. This was 7.22 Km and despite the non line of sight. I was hoping for 2.4 Ghz equipment since it is readily available running 200mw and 24 dbi antennas. I had also planned to mount the units at the antenna point. I was just second guessing the software models results. I am pretty cheap so didn't want to spend the money on equipment if I wasn't reasonably sure it would work. I have never used anything near 2.4 Ghz but I would think it would be rather difficult to line up the antennas. Other than a rough guess what is the best way to do this? It looks like every milliwatt may matter in this installation. Thanks for your help and the explanation of your systems. Tim
