Forum

This information is from teh WFA site.

A mode of operation in which two "channels," or paths on which data can travel, are combined to increase performance in some environments. In the 2.4 GHz frequency band, Wi-Fi CERTIFIED n products are configured to operate using 20 MHz channels by default, and must employ coexistence mechanisms to help ensure that the[b] device defaults to 20 MHz operation when sharing the frequency with other Wi-Fi networks.[/b] In the 5 GHz frequency band, [b]interference is not an issue, so coexistence mechanisms are not required.[/b]
---
How should this be understood?

Does anyone have experience of how those coexistense operatiosn works in addition to the text in the CWNA book?

Does it work with current vendors to use the CTS/RTS feature?

Meru for example allows 40MHz in the 2.4 GHz band, and they can use more than one by turning on 40 MHZ on more than one radio. for example 1+ and 11-.

When turning on 40MHz channels is this optional or mandatory?
Can a client run 40MHz and then drop donw to 20Hz without delays?
Can it be forced to drop down or it is dynamically controlled?
WHich device initiate this?

Reason asking. A lot of g/n devices do not support that many n-features, and may use only g-radios, like the I-phone at max 65 Mbps SISO.

This definition of 40 MHz mode in their glossary is a bit confusing. With all vendors (including Meru) it is a best practice NOT to use 40 MHz channels in 2.4 GHz. In an isolated RF environment where there are no 802.11a/b/g clients, you could use a 40 MHz channel, but you only get one non-overlapping 40 MHz channel. In almost every network environment everywhere, you'll get better performance by just using 20 MHz channels in 2.4 GHz.

In any case, there is no such thing as 40 MHz Wi-Fi operation. There is 20 MHz or 20/40 MHz. Coexistence mechanisms are always required, but the actual mechanisms in use will depend upon the supported features of the product, the composition of all surrounding STAs, the configuration of the AP, the regulatory requirements, etc. 802.11n coexistence has many many layers.

RTS/CTS is one coexistence mechanism that can be used. There is also something called 40 MHz intolerant, which allows an HT client station to notify the AP that the AP should not use 40 MHz channels in an area.

With the iphone, you really shouldn't see a 40 MHz issue since the iphone is 2.4 GHz only. Even if devices like this are only 1x1, it is still helpful for them to be 802.11n. Though they won't see as great of throughput, they at least understand 802.11n HT-mixed format frames, which means that the iphone won't force the AP to use protection (as it would for 802.11a/b/g devices). Avoiding protection can be a big boost to the rest of the cell, assuming no 802.11a/b/g devices are present.

I'd encourage you to take a look at our 802.11n video. There's a lot of good background info in there about protection mechanisms, modes, formats, etc. I'm sure I didn't answer all of your questions here, so feel free to re-ask.

"Meru for example allows 40MHz in the 2.4 GHz band,"
Is this a common feature among AP vendors?

And why didn't the /b ,/g, and especially /a (5GHz) workgroups incorporate (at least as an option) the 40MHz channel? SOHO customers drove the explosive growth of WiFi, and SOHO's usually only need one channel.

This is a guess and only a guess.

Any PHY enhancement (like the MIMO features as well as 40 MHz channels) requires that the MAC can keep up. In other words, if you double the amount of received data at the PHY level by doubling the channel bandwidth, your MAC functionality must be able to process double the data or you get internal congestion and dropped frames on receive or your MAC simply can't supply enough data to the PHY for transmit. So, MAC enhancements (like frame aggregation) would be necessary. It's quite likely that 802.11 chips were not capable of doubling the MAC processing capabilities back in 1999 and early 21st century to keep up with a potential 40 MHz channel for earlier PHYs.

Consider, for example, that many of the early 802.11b devices couldn't even do AES/CCMP. This is evidence that the processing capabilities were limited, and even though better PHY capabilities were within the reach of the EEs designing the equipment, cost-effective 802.11 chips (especially for SOHO markets) that could keep up with the PHY were not possible.

Marcus

Oh, and yes, most vendors allow 40 MHz channels in the 2.4 GHz band, but there's no sense in using them in most cases.