please ask someone else to clarify it for you.
I find it interesting that you
Don't worry Jim, there is no conspiracy here!
I never accused you of conspiracy.
After I read your comments about WISP engineers, "I've delt with far too many of these people. They think they know what they're doing, and they don't," I began to wonder why you have such a bad experience and didn't read the rest of your post closely. Luckily for me, since starting development of WISP systems in 1994, I have had much better experiences.
You sound *just* like the guys at Vistard used to, before we told them to go away back at Vivato. Of course, now all the people at Vivato who claimed that the WISP market for the product wasn't worth the effort are gone, and *presto* Vivato has *hired* the idiots formerly known as Vistard.
Hmm.
Then there was this $2 billion dollar fiasco known as "Project Angel" back at AT&T. They failed in licensed spectrum with a no-cost budget and some of the brightest minds anywhere.
You see, it always comes down to being interference limited. Always.
I'm not sure how any WISP can offer a (real) "service" over wireless in unlicensed spectrum when all I need to do in order to take the "service" offline is to send a signal on the right frequency.
Potentially I can park my car at the base of the tower and take the entire service off-line. (There are other more sophisticated attacks.)
If you think the Internet has security and DOS problems, add wireless.
a) won't re-run your experiment
We had planned to do it with the new eight port cards -- these things take time -- , and today one my techs put together a giant mess of antennas, cables, monitors, routers, and so on.
b) won't quote the current consumption spec of the microtik cards.
The Atheros cards we sell meet the specs at about ~2W (one of them has 2.015w-2.151W as the highest depending on temperature and such -- they have a small linear power converter that loses efficiency in high temperatures).
the linear needs only to convert 3.3V to 2.5V if I recall the Atheros docs correctly. More likely this is your PA's issue.
And is this a max spec, or a 'duty cycle' spec?
SOME TEST RESULTS
-----------------
At this moment, I have connected my laptop (about four six meters from the table that has 5GHz feed horns and it is not pointed my direction) to one of the APs on the eight port card -- I am receiving an average of 26Mb/s. When I switch my laptop to transmit, the AP server receives about 27Mb/s average.
All of the other APs on the server are doing bandwith test with an aggregate of 175Mb/s (data going over the wireless connections). We have some problems with the bandwidth testers -- mostly the software was designed for two connections. So, it appears the eight connections has slowed them down some. We had all eight APs doing 240Mb/s at one time.
Of course, the feed horns (16 of them) are pointing every direction and there is a mess of cables and bandwidth testers connected with gig Ethernet to the routers. I can't say it is perfect -- we will try to organize it better in the next week or two and make a paper showing the results -- with all the settings and screenshots and such.
Just use a a pair of simple omni antennas 2 APs and 2 clients. Put the APs on two different channels (in the same band) and go to distance (at any modulation rate) with the clients. The clients can run 'ftp gets', ttcp, or netperf/iperf. You don't need "bandwidth testers".
I think you'll immediately see the effect I'm referencing.
Then you can move to your horn antennas, etc. (Do your WISPS deploy with horns?)
Was that "four six meters" four to six (4-6) meters, or fourty-six (46) meters?
If you do publish a paper, be prepared to cite the gain of each of the antenna system towards the other antennas. (This should be a measured value, not an estimate), as well as the tested tx power and rx sensitivy of your cards.
Thanks,
Jim
