Slashdot Mirror

I would really like to have that demonstrated to me. My telephone cares little which way I point it.

Your telephone is not an antenna on a cell system. I could have been more explicit by saying "the antenna on a cell system tower", but I assumed most people would know what I was talking about when I referred to plural antennas.

You cell phone is half the connection, and while it is approximately omnidirectional, the antennas at the other, higher power transmitter end, are not.

Maybe the donut is flattened on one axis, but we all know that there aren't tens of thousands of antenna on the tower, each one mounted on independent gimbals, tracking each user.

You missed the critical word "phased". Please refer to here, here, or here for more info. Suffice to say, your hyperbole that one would need tens of thousands of antennas on gimbals to provide directional access to multiple users is quite hilarious.

Your low powered transmitter on your handheld device is not going to be a source of interference to anyone. First, it's low powered. Second, it transmits on the handset side of the frequencies, so even of you are standing next to another cell phone it won't be received by that phone, which is receiving on the tower frequencies. Third, a major design criterion of the cell system is the reuse of frequencies and the allocation of a specific set to cover a certain area. If a StingRay comes to town, it will use a different set of frequencies. It isn't trying to overpower another cell on the same frequencies, it is using its own set at higher power. The phones preferentially connect to the higher power signal. Why? Because it is more likely they can use lower power to communicate and thus save battery life.

I would go with a Yagi or a multi element beam. http://en.wikipedia.org/wiki/Yagi-Uda_antenna
You could start out with just a half wave dipole .
Here a a very good site to get you started.
http://www.antenna-theory.com/

dude, unless your tooth is 250meters in diameter, I'm already talking that.

And no they DONT get a signal well above the noise level, not at 20Ly out.

http://www.antenna-theory.com/basics/friis.php - start there to learn about path losses.

A 10,000 Watt signal into a 70DBm antenna and assuming the receiving side has a sensitivity of -124Dbm Like that of the equipment at Arecebo, I'm not assuming super spaceman unobtanium.

The signal at arrival after only 1 light year will be -161Dbm -170 is considered the noise floor on a super clear frequency in the Ghz range.

This is Basic RF stuff here. Do you not know this?

It does not matter, even if they were 10 lightyears away they can not detect any of our radio signals...

The 50 Watts from a GPS Sattelite that is 20,000Km away will only have about 2.3e-16% of that signal arriving at the antenna. Radio signals don't go on forever at full power, they drop off really fast.

http://www.antenna-theory.com/basics/friis.php The Friis Equation is used to calculate this. and if you calculated out a 1Megawatt TV station transmit power to cover even a small 1 lightyear distance... Unless they are intently looking for the signal, and we transmitted at the quietest frequency in the galaxy (Low background noise) ther estill have a very good chance of never seeing our signal.

So at 20 Lightyears, all hope is lost trying to get any TV or FM radio... the only chance is a tightly beamed intentional ultra high power signal in hopes that the target will be listening in 20 years and they have the same technology and frequency range. You have a better chance of getting hit by lightning from a thunderstorm in china while you are in Iowa.

Next, use channels that have lesser bandwidth consumption. It's not important how many ap's are on one channel, it's important how many data frames are going in and out on that channel.

How do you check that?

I used atheros, madwifi drivers, monitor mode... and tcpdump. Linux of course.
However, mikrotik should be able to do that also. Snooper option - or frequency monitor.

But be aware, when interface is in monitor mode, it only receives packets, so don't start it on a link that transports your packets to the AP.

If you'r using fabric access points (clients), you can't do any of this. Buy an atheros card. It works great under Linux, and you can get pretty cheap one, tp-link, 651 for b/g, or 653 for a.
Linux 2.6.28 introduced athk5 driver, so if you have it, just do modprobe ath5k and you'r ready to go.

It's dB (ratio), not dBm (power level). See http://www.antenna-theory.com/basics/antennapol.php [antenna-theory.com].

You'r right, sorry, my bad. Woked up 5 minutes before writing post.

It's dB (ratio), not dBm (power level). See http://www.antenna-theory.com/basics/antennapol.php.