On Sun, 4 Feb 2007 11:59:18 -0800, Lonnie Nunweiler wrote > I know this goes farther than the B versus G debate that was started, > but the key thing in being able to do this is the cloaking with its > reduced RF spectrum use. A B mode AP cannot do cloaking, nor can > your AP do it if the AP is not an Atheros with a driver that > properly supports the ability.
It must be, because running your gear, I cannot get G mode to work acceptably AT ALL. In my area, every channel has SOME noise on it. Even with signal levels in the low '60's, I could never achieve better than 350 to 400 KB / sec throughput for a DEPLOYED AP and client, and B mode could hit 1400 KB/sec using compressible data, about 650-700 wihtout compression. Narrowing channels appears to kill the G characteristic of waiting for completely clear air before it will transmit. Without cloaking, a nearly idle access point in G mode with a G client, will have varying 1 to 400 ms pings as it waits for clear air to transmit in. Switching to B mode gives you rock solid 1 to 7 ms pings on an active AP with a number of clients. > > B is dead and is holding the Industry back. If you use B mode then > you NEED 400 mW radios because of the noise. Nonsense. My highest power radios are CM9's and I have have few to no noise issues in B mode. B has limited throughput and yet it has it's uses. It is certainly NOT "holding industry back". I believe that investing in B only technology is dumb, though. I thought it was dumb when I started a little less than 3 years ago, which is why I tried not to. I've found that 11a is actually a bit more friendly, in that it's easier to target your ap's and clients, and exclude noise sources outside the pattern. If you use G mode and > X2 cloaking then you need less than 100 mW and you'll have WAY > better performance. Just to be sure about this point --> I am > speaking from EXPERIENCE. This is not some plan I someday hope to > try. It is what we use and is what a lot of others use as well. Sub 100 MW works awesome in B mode, too, so long as the writers of the drivers dont' disable the awesome enhanced features available in Atheros based radios. ( HINT HINT ) > > OFDM was invented as an improvement over previous modulation > techniques. Why do people have such a hard time accepting that it > actually works better? Is it because you have an investment in B > only radios and realize you have to reinvest in G radios? It is > sort of like the phone companies hanging onto their copper lines. Wireless > started to cream them and now you are seeing that G is creaming B, so > that the old established operators are in trouble. That's a lotta hype. I put YOUR gear in place, as per YOUR instructions, and YOUR predictions don't work out that way. I've found that there's caveats to all this. OFDM makes great RF links, but it takes a little bit more signal to maintain low retransmissions or errors. On the other hand, OFDM is dramatically better when it comes to surviving multipath issues and fresnel encroachment. G mode is simply not workable in a busy environment, unless you can force the radio to abandon listening to non ofdm noise, or narrow your channels enough to get away from it. By design, standard 11g can have it's performance killed by even a single B client attempting to associate to the AP. Not explaining this to people wanting to implement is irresponsible, in my view. > > Lonnie > > -- > Lonnie Nunweiler > Valemount Networks Corporation > http://www.star-os.com/ > -- > WISPA Wireless List: wireless@wispa.org > > Subscribe/Unsubscribe: > http://lists.wispa.org/mailman/listinfo/wireless > > Archives: http://lists.wispa.org/pipermail/wireless/ -------------------------------------------- Mark Koskenmaki <> Neofast, Inc Broadband for the Walla Walla Valley and Blue Mountains 541-969-8200 -- WISPA Wireless List: wireless@wispa.org Subscribe/Unsubscribe: http://lists.wispa.org/mailman/listinfo/wireless Archives: http://lists.wispa.org/pipermail/wireless/