Posted: Fri Jul 14, 2006 15:58 Post subject: wrt54gl transmit power
hello. av got 4 linksys wrt54gl routers which i'm using in a WDS mode. they're all within a radius of 150m. i've hacked d firmware adn they're all running DD-WRTv.23. my transmit power is set at 70mW as recommended. pls i'd like to konw wat transmit settings are optimal adn would cause minimum interference. any1 who konws? tanx
Joined: 07 Jun 2006 Posts: 94 Location: North Idaho (USA)
Posted: Mon Jul 17, 2006 23:52 Post subject: Re: wrt54gl transmit power
myroman wrote:
hello. av got 4 linksys wrt54gl routers which i'm using in a WDS mode. they're all within a radius of 150m. i've hacked d firmware adn they're all running DD-WRTv.23. my transmit power is set at 70mW as recommended. pls i'd like to konw wat transmit settings are optimal adn would cause minimum interference. any1 who konws? tanx
The optimal transmit power (for clients and AP radios) is the lowest transmit power that works in your network. A transmit power higher than needed just creates more noise and interefernce for other wireless devices.
Note - A high power AP really does no good if all the clients are notebook computers transmiting at 28 mw.
Good external antennas are the best solution for any radio network. With good antennas, you can actually run very low power. I have a client about 150 meters away from my AP where the client radio is set to only 9 mw (19 dBi antenna) and they connect at 54 Meg. _________________ ******************************************
Tom Jones
North Idaho - USA
******************************************
The low level firmware on the radio card ignores power setting attempts above the level that the vendor used for FCC certification and for WiFi compliance proof. Running anything higher in the 11g modes distorts the OFDM signal and causes higher error rates at the receiver. The measure of this is the "EVM" - explained in the 802.11 specs.
And there's the imbalanced power issue as above- it does no good to have a boomer access point/w-router if the clients remain peanut whistles. Like a speaker with a Public Address system can't hear the unaided voices in the back of the auditorim.
That, unfortunatly, is overly simplistic. What if the AP has better recieve sensitivity than the client (which is almost certainly the case.) In that case it is not true that it doesn't help to have the AP at a higher power than the nic.
Think about it this way: do you really think that cell phone towers and cell phones tansmit at the same strength?
Ill tell you what you will not see and improvment in dbi on the unit that you incrase the mw on you will se it on the client I know i have done this with my WDS routers they are about 450 feet away When they are both increased to 70 it improves both when one is it improves the signal of the opposing router if you want a cheap solution to double your power go hear
That, unfortunatly, is overly simplistic. What if the AP has better recieve sensitivity than the client (which is almost certainly the case.) In that case it is not true that it doesn't help to have the AP at a higher power than the nic.
Think about it this way: do you really think that cell phone towers and cell phones tansmit at the same strength?
indeed, the radio chipsets in the access points and clients are essentially the same; same power, same receiver sensitivity. Access points might have 2-3 dBi gain antennas whereas a client radio as in laptops has little or negative gain and a sub-par antenna pattern.
as to cell phones - well, cell phones are not CSMA like WiFi; they are full duplex and use two widely separate frequncies - uplink and downlink. As you may know, CDMA (Verizon,Sprint) works on the principle that in both directions, every packet to every subscriber is transmitted at just enough power so it can be received. In CDMA, keeping the power low (it's adjusted hundreds of times per second) is how it gets higher capacity per channel - like 30 simultaneous calls in 1.25MHz. In TDMA/GSM (all the other phone companies), and iDEN (Nextel), the power control is not as critical and efficient.
So, for WiFi, the 11g chipsets (there are but 4 or so vendors) are certified for the FCC then for WiFi with a specific EVM. Because they don't want to violate, the radios don't really operate at much more than what they were certified at- the power amplifiers aren't linear enough to do so. No matter what the HTTP GUI may say.
