Our primary test site turned out to be inaccessible, so we had to
fallback to another test site. We ended up NOT having clear line of
sight over the 7.23 mile shot. In fact, about 900 meters of the shot
went through the tops of evergreen trees. That makes this test invalid.
Having said that, the received signal strength from the +10dBm
transmitter was about -100dBm, which is 40dB below projected values. So
about 105dB of path loss if you discount the cable losses.
We then hooked up the modems which run +30dBm and they were able to
establish a 39Mbit air speed link. Link quality was about 60% both
ways. Throughput tests showed 30Mbit(min) to 33Mbit(max) data
transmission rate. Speeds were the same in both directions. It was
nice and steady, with good ping times throughout. Very well behaved
despite the harsh conditions. The radios reported -75dBm signal. The
math predicts -85dBm should have been reported. The radios are not
calibrated measurement devices though, so a 10dB error is not surprising.
This whole test needs to be re-done with actual line-of-sight conditions
to get a good reading on path loss in free space. A couple sites look
possible on satellite but need to be verified on the ground for
suitability. They're only 6.9 and 6.7 miles long though. It's tough
getting good clear long distance shots in publicly available places. If
anyone wants to volunteer to help out with follow-up tests, let me
know. If you know of a good long distance LoS option, also let me know.
--Bart
